Professional Documents
Culture Documents
Electromagnetic Field Theory - Bo Thide
Electromagnetic Field Theory - Bo Thide
Electromagnetic Field Theory - Bo Thide
MAGNETIC
FIELD
THEORY
Bo Thid
P S I L O N
E D I A
Bo Thid
Also available
E LECTROMAGNETIC
F IELD T HEORY
Bo Thid
Swedish Institute of Space Physics
and
Department of Astronomy and Space Physics
Uppsala University, Sweden
UPSILON
MEDIA
UPPSALA
SWEDEN
Contents
Preface
xi
1 Classical Electrodynamics
1.1 Electrostatics . . . . . . . . . . . . . . .
1.1.1 Coulombs law . . . . . . . . . .
1.1.2 The electrostatic field . . . . . . .
1.2 Magnetostatics . . . . . . . . . . . . . .
1.2.1 Ampres law . . . . . . . . . . .
1.2.2 The magnetostatic field . . . . . .
1.3 Electrodynamics . . . . . . . . . . . . .
1.3.1 Equation of continuity . . . . . .
1.3.2 Maxwells displacement current .
1.3.3 Electromotive force . . . . . . . .
1.3.4 Faradays law of induction . . . .
1.3.5 Maxwells microscopic equations
1.3.6 Maxwells macroscopic equations
1.4 Electromagnetic Duality . . . . . . . . .
1
1
1
2
4
4
6
8
9
9
10
11
14
14
15
Example 1.1 Duality of the electromagnetodynamic equations 16
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2 Electromagnetic Waves
2.1 The wave equation . . . . . . . . . . . . . . . . .
2.1.1 The wave equation for E . . . . . . . . . .
2.1.2 The wave equation for B . . . . . . . . . .
2.1.3 The time-independent wave equation for E
2.2 Plane waves . . . . . . . . . . . . . . . . . . . . .
2.2.1 Telegraphers equation . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
20
23
24
24
24
25
26
27
ii
C ONTENTS
29
30
31
Electromagnetic Potentials
3.1 The electrostatic scalar potential . . . . . . . . . . . . . . . .
3.2 The magnetostatic vector potential . . . . . . . . . . . . . . .
3.3 The electromagnetic scalar and vector potentials . . . . . . . .
3.3.1 Electromagnetic gauges . . . . . . . . . . . . . . . .
Lorentz equations for the electromagnetic potentials .
Gauge transformations . . . . . . . . . . . . . . . . .
3.3.2 Solution of the Lorentz equations for the electromagnetic potentials . . . . . . . . . . . . . . . . . . . . .
The retarded potentials . . . . . . . . . . . . . . . . .
Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
33
33
34
34
36
36
36
43
45
47
49
Relativistic Electrodynamics
5.1 The special theory of relativity . . . . . . . . . . . . . . . . .
5.1.1 The Lorentz transformation . . . . . . . . . . . . . .
5.1.2 Lorentz space . . . . . . . . . . . . . . . . . . . . . .
Metric tensor . . . . . . . . . . . . . . . . . . . . . .
Radius four-vector in contravariant and covariant form
Scalar product and norm . . . . . . . . . . . . . . . .
Invariant line element and proper time . . . . . . . . .
Four-vector fields . . . . . . . . . . . . . . . . . . . .
The Lorentz transformation matrix . . . . . . . . . . .
The Lorentz group . . . . . . . . . . . . . . . . . . .
5.1.3 Minkowski space . . . . . . . . . . . . . . . . . . . .
5.2 Covariant classical mechanics . . . . . . . . . . . . . . . . .
5.3 Covariant classical electrodynamics . . . . . . . . . . . . . .
5.3.1 The four-potential . . . . . . . . . . . . . . . . . . .
5.3.2 The Linard-Wiechert potentials . . . . . . . . . . . .
5.3.3 The electromagnetic field tensor . . . . . . . . . . . .
Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
51
51
52
53
54
54
55
56
57
57
58
58
61
62
62
63
65
67
38
41
41
iii
Other fields . . . . . . . . . . . . . . . . . . . . . . .
Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
7 Interactions of Fields and Matter
7.1 Electric polarisation and the electric displacement vector
7.1.1 Electric multipole moments . . . . . . . . . . .
7.2 Magnetisation and the magnetising field . . . . . . . . .
7.3 Energy and momentum . . . . . . . . . . . . . . . . . .
7.3.1 The energy theorem in Maxwells theory . . . .
7.3.2 The momentum theorem in Maxwells theory . .
Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
8 Electromagnetic Radiation
8.1 The radiation fields . . . . . . . . . . . . . . . . . .
8.2 Radiated energy . . . . . . . . . . . . . . . . . . . .
8.2.1 Monochromatic signals . . . . . . . . . . . .
8.2.2 Finite bandwidth signals . . . . . . . . . . .
8.3 Radiation from extended sources . . . . . . . . . . .
8.3.1 Linear antenna . . . . . . . . . . . . . . . .
8.4 Multipole radiation . . . . . . . . . . . . . . . . . .
8.4.1 The Hertz potential . . . . . . . . . . . . . .
8.4.2 Electric dipole radiation . . . . . . . . . . .
8.4.3 Magnetic dipole radiation . . . . . . . . . .
8.4.4 Electric quadrupole radiation . . . . . . . . .
8.5 Radiation from a localised charge in arbitrary motion
8.5.1 The Linard-Wiechert potentials . . . . . . .
8.5.2 Radiation from an accelerated point charge .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
69
69
69
69
72
76
77
80
81
84
85
.
.
.
.
.
.
.
87
87
87
90
91
92
93
95
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Example 8.1 The fields from a uniformly moving charge .
97
97
99
100
100
102
102
104
104
108
109
110
111
112
114
121
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
iv
C ONTENTS
8.5.3
8.5.4
Vavilov-Cerenkov
radiation . . . . . . .
Bibliography . . . . . . . . . . . . . . . . . . . . . .
.
.
.
.
.
.
.
.
132
134
134
137
137
139
142
147
F Formulae
F.1 The Electromagnetic Field . . . . . . . . . . . . . . . . . . .
F.1.1 Maxwells equations . . . . . . . . . . . . . . . . . .
Constitutive relations . . . . . . . . . . . . . . . . . .
F.1.2 Fields and potentials . . . . . . . . . . . . . . . . . .
Vector and scalar potentials . . . . . . . . . . . . . .
Lorentz gauge condition in vacuum . . . . . . . . . .
F.1.3 Force and energy . . . . . . . . . . . . . . . . . . . .
Poyntings vector . . . . . . . . . . . . . . . . . . . .
Maxwells stress tensor . . . . . . . . . . . . . . . . .
F.2 Electromagnetic Radiation . . . . . . . . . . . . . . . . . . .
F.2.1 Relationship between the field vectors in a plane wave
F.2.2 The far fields from an extended source distribution . .
F.2.3 The far fields from an electric dipole . . . . . . . . . .
F.2.4 The far fields from a magnetic dipole . . . . . . . . .
F.2.5 The far fields from an electric quadrupole . . . . . . .
F.2.6 The fields from a point charge in arbitrary motion . . .
F.2.7 The fields from a point charge in uniform motion . . .
F.3 Special Relativity . . . . . . . . . . . . . . . . . . . . . . . .
F.3.1 Metric tensor . . . . . . . . . . . . . . . . . . . . . .
F.3.2 Covariant and contravariant four-vectors . . . . . . . .
F.3.3 Lorentz transformation of a four-vector . . . . . . . .
F.3.4 Invariant line element . . . . . . . . . . . . . . . . . .
F.3.5 Four-velocity . . . . . . . . . . . . . . . . . . . . . .
F.3.6 Four-momentum . . . . . . . . . . . . . . . . . . . .
F.3.7 Four-current density . . . . . . . . . . . . . . . . . .
F.3.8 Four-potential . . . . . . . . . . . . . . . . . . . . . .
149
149
149
149
149
149
150
150
150
150
150
150
150
150
151
151
151
151
152
152
152
152
152
152
153
153
153
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
153
153
154
154
154
154
154
154
154
154
156
157
157
Appendices
148
M Mathematical Methods
M.1 Scalars, Vectors and Tensors
M.1.1 Vectors . . . . . . .
Radius vector . . . .
M.1.2 Fields . . . . . . . .
Scalar fields . . . . .
Vector fields . . . .
Tensor fields . . . .
159
159
159
159
161
161
161
162
164
167
167
167
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Example M.1 Tensors in 3D space .
M.1.3 Vector algebra . . . . . . . . . .
Scalar product . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Example M.2 Inner products in complex vector space .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Dyadic product .
Vector product .
M.1.4 Vector analysis .
The del operator
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Example M.5 The four-del operator in Lorentz space .
The gradient . . . . . . . . . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
169
170
170
170
171
172
vi
C ONTENTS
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
173
173
173
174
174
175
176
176
176
177
List of Figures
1.1
1.2
1.3
Coulomb interaction . . . . . . . . . . . . . . . . . . . . . .
Ampre interaction . . . . . . . . . . . . . . . . . . . . . . .
Moving loop in a varying B field . . . . . . . . . . . . . . . .
2
5
12
5.1
5.2
5.3
52
59
59
6.1
76
8.1
8.2
8.3
8.4
8.5
8.6
8.7
8.8
8.9
Vavilov-Cerenkov
cone . . . . . . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
98
112
114
128
129
133
135
138
144
vii
To the memory of
L EV M IKHAILOVICH E RUKHIMOV
dear friend, remarkable physicist
and a truly great human.
Preface
This book is the result of a twenty-five year long love affair. In 1972, I took
my first advanced course in electrodynamics at the Theoretical Physics department, Uppsala University. Shortly thereafter, I joined the research group there
and took on the task of helping my supervisor, professor P ER -O LOF F R MAN , with the preparation of a new version of his lecture notes on Electricity
Theory. These two things opened up my eyes for the beauty and intricacy of
electrodynamics, already at the classical level, and I fell in love with it.
Ever since that time, I have off and on had reason to return to electrodynamics, both in my studies, research and teaching, and the current book
is the result of my own teaching of a course in advanced electrodynamics at
Uppsala University some twenty odd years after I experienced the first encounter with this subject. The book is the outgrowth of the lecture notes that I
prepared for the four-credit course Electrodynamics that was introduced in the
Uppsala University curriculum in 1992, to become the five-credit course Classical Electrodynamics in 1997. To some extent, parts of these notes were based
on lecture notes prepared, in Swedish, by B ENGT L UNDBORG who created,
developed and taught the earlier, two-credit course Electromagnetic Radiation
at our faculty.
Intended primarily as a textbook for physics students at the advanced undergraduate or beginning graduate level, I hope the book may be useful for
research workers too. It provides a thorough treatment of the theory of electrodynamics, mainly from a classical field theoretical point of view, and includes such things as electrostatics and magnetostatics and their unification
into electrodynamics, the electromagnetic potentials, gauge transformations,
covariant formulation of classical electrodynamics, force, momentum and energy of the electromagnetic field, radiation and scattering phenomena, electromagnetic waves and their propagation in vacuum and in media, and covariant
Lagrangian/Hamiltonian field theoretical methods for electromagnetic fields,
particles and interactions. The aim has been to write a book that can serve
both as an advanced text in Classical Electrodynamics and as a preparation for
studies in Quantum Electrodynamics and related subjects.
In an attempt to encourage participation by other scientists and students in
xi
xii
P REFACE
the authoring of this book, and to ensure its quality and scope to make it useful
in higher university education anywhere in the world, it was produced within
a World-Wide Web (WWW) project. This turned out to be a rather successful
move. By making an electronic version of the book freely down-loadable on
the net, I have not only received comments on it from fellow Internet physicists
around the world, but know, from WWW hit statistics that at the time of
writing this, the book serves as a frequently used Internet resource. This way
it is my hope that it will be particularly useful for students and researchers
working under financial or other circumstances that make it difficult to procure
a printed copy of the book.
I am grateful not only to Per-Olof Frman and Bengt Lundborg for providing the inspiration for my writing this book, but also to C HRISTER WAHLBERG
at Uppsala University for interesting discussions on electrodynamics in general
and on this book in particular, and to my former graduate students M ATTIAS
WALDENVIK and T OBIA C AROZZI as well as A NDERS E RIKSSON, all at the
Swedish Institute of Space Physics, Uppsala Division, and who have participated in the teaching and commented on the material covered in the course
and in this book. Thanks are also due to my long-term space physics colleague H ELMUT KOPKA of the Max-Planck-Institut fr Aeronomie, Lindau,
Germany, who not only taught me about the practical aspects of the of highpower radio wave transmitters and transmission lines, but also about the more
delicate aspects of typesetting a book in TEX and LATEX. I am particularly
indebted to Academician professor V ITALIY L. G INZBURG for his many fascinating and very elucidating lectures, comments and historical footnotes on
electromagnetic radiation while cruising on the Volga river during our joint
Russian-Swedish summer schools.
Finally, I would like to thank all students and Internet users who have
downloaded and commented on the book during its life on the World-Wide
Web.
I dedicate this book to my son M ATTIAS, my daughter K AROLINA, my
high-school physics teacher, S TAFFAN RSBY, and to my fellow members of
the C APELLA P EDAGOGICA U PSALIENSIS.
Uppsala, Sweden
November, 2000
B O T HID
1
Classical
Electrodynamics
Classical electrodynamics deals with electric and magnetic fields and interactions caused by macroscopic distributions of electric charges and currents.
This means that the concepts of localised charges and currents assume the
validity of certain mathematical limiting processes in which it is considered
possible for the charge and current distributions to be localised in infinitesimally small volumes of space. Clearly, this is in contradiction to electromagnetism on a truly microscopic scale, where charges and currents are known to
be spatially extended objects. However, the limiting processes used will yield
results which are correct on small as well as large macroscopic scales.
In this Chapter we start with the force interactions in classical electrostatics and classical magnetostatics and introduce the static electric and magnetic
fields and find two uncoupled systems of equations for them. Then we see how
the conservation of electric charge and its relation to electric current leads to
the dynamic connection between electricity and magnetism and how the two
can be unified in one theory, classical electrodynamics, described by one system of coupled dynamic field equations.
1.1 Electrostatics
The theory that describes physical phenomena related to the interaction between
stationary electric charges or charge distributions in space is called electrostatics.
C LASSICAL E LECTRODYNAMICS
x x0
q0
x0
O
F IGURE 1.1: Coulombs law describes how a static electric charge q,
located at a point x relative to the origin O, experiences an electrostatic
force from a static electric charge q0 located at x0 .
Let F denote the force acting on a charged particle with charge q located at x,
due to the presence of a charge q0 located at x0 . According to Coulombs law
this force is, in vacuum, given by the expression
qq0
1
qq0 x x0
=
F(x) =
3
0
40 |x x |
40
|x x0 |
(1.1)
where we have used results from Example M.6 on page 172. In SI units, which
we shall use throughout, the force F is measured in Newton (N), the charges
q and q0 in Coulomb (C) [= Ampre-seconds (As)], and the length |x x0 | in
metres (m). The constant 0 = 107 /(4c2 ) 8.8542 1012 Farad per metre
(F/m) is the vacuum permittivity and c 2.9979 108 m/s is the speed of light
in vacuum. In CGS units 0 = 1/(4) and the force is measured in dyne, the
charge in statcoulomb, and length in centimetres (cm).
Estat lim
q0
F
q
(1.2)
1.1
E LECTROSTATICS
purpose of the limiting process is to assure that the test charge q does not
influence the field, the expression for Estat does not depend explicitly on q but
only on the charge q0 and the relative radius vector x x0 . This means that we
can say that any net electric charge produces an electric field in the space that
surrounds it, regardless of the existence of a second charge anywhere in this
space.1
Using formulae (1.1) and (1.2), we find that the electrostatic field Estat at
the field point x (also known as the observation point), due to a field-producing
charge q0 at the source point x0 , is given by
stat
1
q0 x x 0
q0
(x) =
40 |x x0 |3
40
|x x0 |
(1.3)
q0i x x0i
40 x x0 3
i
(1.4)
If the discrete charges are small and numerous enough, we introduce the charge
density located at x0 and write the total field as
E
stat
1
(x) =
40
1
x x0 3 0
d x =
(x )
3
0
4
x
|x
|
V
0
0
(x )
V
1
d3x0 (1.5)
|x x0 |
where, in the last step, we used formula Equation (M.68) on page 172. We
emphasise that Equation (1.5) above is valid for an arbitrary distribution of
charges, including discrete charges, in which case can be expressed in terms
of one or more Dirac delta functions.
1 In the preface to the first edition of the first volume of his book A Treatise on Electricity
and Magnetism, first published in 1873, James Clerk Maxwell describes this in the following,
almost poetic, manner: [6]
For instance, Faraday, in his minds eye, saw lines of force traversing all space
where the mathematicians saw centres of force attracting at a distance: Faraday
saw a medium where they saw nothing but distance: Faraday sought the seat of
the phenomena in real actions going on in the medium, they were satisfied that
they had found it in a power of action at a distance impressed on the electric
fluids.
2 In
C LASSICAL E LECTRODYNAMICS
stat
1
1
(x0 )
d3x0
(x) =
40
|x x0 |
V
1
1
0
(x )
=
d3x0
40 V
|x x0 |
=0
(1.6)
x
40 V
|x |
1
1
(x0 )2
=
d3x0
40 V
|x x0 |
1
=
(x0 )(x x0 ) d3x0
0 V
(x)
=
0
Estat (x) =
(1.7)
1.2 Magnetostatics
While electrostatics deals with static charges, magnetostatics deals with stationary currents, i.e., charges moving with constant speeds, and the interaction
between these currents.
1.2
M AGNETOSTATICS
J
C
dl
x x0
dl0
x
C0
x0
J0
O
F IGURE 1.2: Ampres law describes how a small loop C, carrying a
static electric current J through its tangential line element dl located at
x, experiences a magnetostatic force from a small loop C 0 , carrying a
static electric current J 0 through the tangential line element dl0 located at
x0 . The loops can have arbitrary shapes as long as they are simple and
closed.
located at x0 . According to Ampres law this force is, in vacuum, given by the
expression
0 JJ 0
dl0 (x x0 )
dl
4 C C 0
|x x0 |3
0 JJ 0
1
0
=
dl dl
4 C C 0
|x x0 |
F(x) =
(1.8)
Here dl and dl0 are tangential line elements of the loops C and C 0 , respectively,
and, in SI units, 0 = 4 107 1.2566 106 H/m is the vacuum permeability. From the definition of 0 and 0 (in SI units) we observe that
0 0 =
107
1
(F/m) 4 107 (H/m) = 2 (s2 /m2 )
2
4c
c
(1.9)
C LASSICAL E LECTRODYNAMICS
formula (F.54) on page 155, we can rewrite (1.8) in the following way
1
0 JJ 0
F(x) =
dl
dl0
4 C C 0
|x x0 |
x x0
0 JJ 0
dl dl0
4 C C 0 |x x0 |3
(1.10)
Recognising the fact that the integrand in the first integral is an exact differential so that this integral vanishes, we can rewrite the force expression, Equation (1.8) on the previous page, in the following symmetric way
0 JJ 0
4
F(x) =
C0
x x0
dl dl0
|x x0 |3
(1.11)
dBstat (x)
0 J 0 0 x x 0
dl
4
|x x0 |3
(1.12)
which expresses the small element dBstat (x) of the static magnetic field set
up at the field point x by a small line element dl0 of stationary current J 0 at
the source point x0 . The SI unit for the magnetic field, sometimes called the
magnetic flux density or magnetic induction, is Tesla (T).
If we generalise expression (1.12) to an integrated steady state current distribution j(x), we obtain Biot-Savarts law:
0
x x0 3 0
j(x0 )
dx
4 V
|x x0 |3
1
0
d3x0
j(x0 )
=
4 V
|x x0 |
Bstat (x) =
(1.13)
Comparing Equation (1.5) on page 3 with Equation (1.13), we see that there exists a close analogy between the expressions for Estat and Bstat but that they differ in their vectorial characteristics. With this definition of Bstat , Equation (1.8)
on the previous page may we written
F(x) = J
dl Bstat (x)
(1.14)
1.2
M AGNETOSTATICS
stat
0
1
(x) = j(x0 )
d3x0
4 V
|x x0 |
0
1
=
[ j(x0 )]d3x0
4 V
|x x0 |
1
0
0
j(x )
+
d3x0
4 V
|x x0 |
=0
(1.15)
stat
1
0
d3x0
(x) = j(x0 )
0|
x
4
|x
V
0
1
0 2
=
d3x0
j(x )
4 V
|x x0 |
0
1
0
0
0
d3x0
+
[j(x ) ]
4 V
|x x0 |
(1.16)
In the first of the two integrals on the right hand side, we use the representation
of the Dirac delta function Equation (M.73) on page 174, and integrate the
second one by parts, by utilising formula (F.59) on page 155 as follows:
1
d3x0
|x x0 |
1
1
0
3 0
0
0
0
0
= x k j(x ) 0
dx
j(x )
d3x0
0
xk |x x |
|x x0 |
V
V
1
1
0
0
0
0
= x k j(x ) 0
dS
j(x )
d3x0
0|
0|
x
x
|x
|x
S
V
k
(1.17)
0
[j(x ) ]
Then we note that the first integral in the result, obtained by applying Gausss
theorem, vanishes when integrated over a large sphere far away from the localised source j(x0 ), and that the second integral vanishes because j = 0 for
C LASSICAL E LECTRODYNAMICS
stationary currents (no charge accumulation in space). The net result is simply
Bstat (x) = 0
(1.18)
1.3 Electrodynamics
As we saw in the previous sections, the laws of electrostatics and magnetostatics can be summarised in two pairs of time-independent, uncoupled vector
differential equations, namely the equations of classical electrostatics 3
(x)
0
stat
E (x) = 0
Estat (x) =
(1.19a)
(1.19b)
stat
(1.20a)
(x) = 0 j(x)
(1.20b)
Since there is nothing a priori which connects Estat directly with Bstat , we must
consider classical electrostatics and classical magnetostatics as two independent theories.
However, when we include time-dependence, these theories are unified
into one theory, classical electrodynamics. This unification of the theories of
electricity and magnetism is motivated by two empirically established facts:
1. Electric charge is a conserved quantity and current is a transport of electric charge. This fact manifests itself in the equation of continuity and,
as a consequence, in Maxwells displacement current.
2. A change in the magnetic flux through a loop will induce an EMF electric field in the loop. This is the celebrated Faradays law of induction.
3 The
1.3
E LECTRODYNAMICS
= 0 j(t, x) + 0 0 E(t, x)
t
0
3 0
(t, x )
V
1
d3x0
|x x0 |
(1.22)
where, in the last step, we have assumed that a generalisation of Equation (1.5)
on page 3 to time-varying fields allows us to make the identification
1
40 t
1
d3x0 =
0
t
40
|x x |
= E(t, x)
t
(t, x )
V
(t, x )
V
1
d3x0
|x x0 |
(1.23)
Later, we will need to consider this formal result further. The result is Maxwells source equation for the B field
B(t, x) = 0 j(t, x) +
0 E(t, x)
t
(1.24)
10
C LASSICAL E LECTRODYNAMICS
where the last term 0 E(t, x)/t is the famous displacement current. This
term was introduced, in a stroke of genius, by Maxwell in order to make the
right hand side of this equation divergence free when j(t, x) is assumed to represent the density of the total electric current, which can be split up in ordinary conduction currents, polarisation currents and magnetisation currents.
The displacement current is an extra term which behaves like a current density
flowing in vacuum. As we shall see later, its existence has far-reaching physical consequences as it predicts the existence of electromagnetic radiation that
can carry energy and momentum over very long distances, even in vacuum.
(1.25)
where is the electric conductivity (S/m). In the most general cases, for instance in an anisotropic conductor, is a tensor.
We can view Ohms law, Equation (1.25) above, as the first term in a Taylor
expansion of the law j[E(t, x)]. This general law incorporates non-linear effects
such as frequency mixing. Examples of media which are highly non-linear are
semiconductors and plasma. We draw the attention to the fact that even in cases
when the linear relation between E and j is a good approximation, we still have
to use Ohms law with care. The conductivity is, in general, time-dependent
(temporal dispersive media) but then it is often the case that Equation (1.25) is
valid for each individual Fourier component of the field.
If the current is caused by an applied electric field E(t, x), this electric field
will exert work on the charges in the medium and, unless the medium is superconducting, there will be some energy loss. The rate at which this energy is
expended is j E per unit volume. If E is irrotational (conservative), j will
decay away with time. Stationary currents therefore require that an electric
field which corresponds to an electromotive force (EMF) is present. In the
presence of such a field EEMF , Ohms law, Equation (1.25) above, takes the
form
j = (Estat + EEMF )
(1.26)
1.3
11
E LECTRODYNAMICS
C (E
stat
+ EEMF ) dl
(1.27)
CE
EMF
dl
(1.28)
d
m (t, x)
dt
d
=
B(t, x) dS =
dt S
E(t, x) dl =
dS
B(t, x)
t
(1.29)
where m is the magnetic flux and S is the surface encircled by C which can be
interpreted as a generic stationary loop and not necessarily as a conducting
circuit. Application of Stokes theorem on this integral equation, transforms it
into the differential equation
E(t, x) =
B(t, x)
t
(1.30)
which is valid for arbitrary variations in the fields and constitutes the Maxwell
equation which explicitly connects electricity with magnetism.
Any change of the magnetic flux m will induce an EMF. Let us therefore
consider the case, illustrated if Figure 1.3.4 on the following page, that the
loop is moved in such a way that it links a magnetic field which varies during
the movement. The convective derivative is evaluated according to the wellknown operator formula
12
C LASSICAL E LECTRODYNAMICS
dS
B(x)
v
C
dl
B(x)
= +v
(1.31)
dt t
which follows immediately from the rules of differentiation of an arbitrary
differentiable function f (t, x(t)). Applying this rule to Faradays law, Equation (1.29) on the previous page, we obtain
d
B dS = dS B (v )B dS
(1.32)
dt S
S t S
During spatial differentiation v is to be considered as constant, and Equation (1.15) on page 7 holds also for time-varying fields:
E(t, x) =
B(t, x) = 0
(1.33)
(1.34)
1.3
13
E LECTRODYNAMICS
allowing us to rewrite Equation (1.32) on the facing page in the following way:
E(t, x) =
d
dt
EEMF dl =
S
B dS
t
B dS
(1.35)
(B v) dS
EEMF dl =
B dS
t
B dS
t
(B v) dl
(1.36)
(EEMF v B) dl =
(1.37)
where EEMF is the field which is induced in the loop, i.e., in the moving
system. The use of Stokes theorem backwards on Equation (1.37) above
yields
(EEMF v B) =
B
t
(1.38)
(1.39)
(1.40)
(1.41)
Hence, we can conclude that for a stationary observer, the Maxwell equation
E =
B
t
(1.42)
14
C LASSICAL E LECTRODYNAMICS
(t, x)
0
B
=
t
B = 0
E
B 0 0
= 0 j(t, x)
t
E+
(1.43a)
(1.43b)
(1.43c)
(1.43d)
In these equations (t, x) represents the total, possibly both time and space dependent, electric charge, i.e., free as well as induced (polarisation) charges,
and j(t, x) represents the total, possibly both time and space dependent, electric current, i.e., conduction currents (motion of free charges) as well as all
atomistic (polarisation, magnetisation) currents. As they stand, the equations
therefore incorporate the classical interaction between all electric charges and
currents in the system and are called Maxwells microscopic equations. Another name often used for them is the Maxwell-Lorentz equations. Together
with the appropriate constitutive relations, which relate and j to the fields,
and the initial and boundary conditions pertinent to the physical situation at
hand, they form a system of well-posed partial differential equations which
completely determine E and B.
(1.44a)
H = H[t, x; E, B]
(1.44b)
1.4
15
E LECTROMAGNETIC D UALITY
Under certain conditions, for instance for very low field strengths, we may
assume that the response of a substance is linear so that
D = (t, x)E
(1.45)
H = (t, x)B
(1.46)
i.e., that the derived fields are linearly proportional to the primary fields and
that the electric displacement (magnetising field) is only dependent on the electric (magnetic) field.
The field equations expressed in terms of the derived field quantities D and
H are
D = (t, x)
B
=0
E+
t
B = 0
D
H
= j(t, x)
t
(1.47a)
(1.47b)
(1.47c)
(1.47d)
e
0
B
= 0 jm
t
B = 0 m
E
B + 0 0
= 0 je
t
E+
(1.48a)
(1.48b)
(1.48c)
(1.48d)
We shall call these equations the Dirac-Maxwell equations or the electromagnetodynamic equations
16
C LASSICAL E LECTRODYNAMICS
( B) 0 jm 0
t
(1.49)
where we used the fact that, according to formula (M.82) on page 175, the
divergence of a curl always vanishes. Using (1.48c) to rewrite this relation, we
obtain the equation of continuity for magnetic monopoles
m
+ jm = 0
t
(1.50)
which has the same form as that for the electric monopoles (electric charges)
and currents, Equation (1.21) on page 9.
We notice that the new Equations (1.48) on the preceding page exhibit the
following symmetry (recall that 0 0 = 1/c2 ):
E cB
(1.51a)
cB E
(1.51b)
m ce
(1.51d)
jm cje
(1.51f)
ce m
cje jm
(1.51c)
(1.51e)
which is a particular case ( = /2) of the general duality transformation (depicted by the Hodge star operator)
?
(1.52a)
c B = E sin + cB cos
(1.52b)
c e = ce cos + m sin
(1.52c)
(1.52d)
E = E cos + cB sin
m = ce sin + m cos
(1.52e)
(1.52f)
which leaves the Dirac-Maxwell equations, and hence the physics they describe (often referred to as electromagnetodynamics), invariant. Since E and j e
are (true or polar) vectors, B a pseudovector (axial vector), e a (true) scalar,
then m and , which behaves as a mixing angle in a two-dimensional charge
space, must be pseudoscalars and jm a pseudovector.
1.4
17
E LECTROMAGNETIC D UALITY
D UALITY OF
E XAMPLE
1.1
?E +
E sin + B cos
= (E cos + cB sin ) +
t
t c
1 E
B
cos 0 jm cos +
sin + c0 je sin
=
t
c t
1 E
B
sin +
cos = 0 jm cos + c0 je sin
c t
t
= 0 (cje sin + jm cos ) = 0 ?jm
?E = (E cos + cB sin ) =
(1.53)
(1.54)
QED
EXAMPLE
1.1
(1.55a)
(1.55b)
E XAMPLE
1.2
18
C LASSICAL E LECTRODYNAMICS
yields
E = ?E cos c?B sin
(1.57)
(1.58)
Furthermore, from the expressions for the transformed charges and currents above, we
find that
?
1 e
e
?E =
=
(1.59)
0
cos 0
and
?B = 0 ?m = 0
(1.60)
so that
E =
e
1 e
cos 0 =
cos 0
0
(1.61)
QED
OF EXAMPLE
1.2
The invariance of the Dirac-Maxwell equations under the similarity transformation means that the amount of magnetic monopole density m is irrelevant for the physics as long as the ratio m /e = tan is kept constant. So
whether we assume that the particles are only electrically charged or have also
a magnetic charge with a given, fixed ratio between the two types of charges
is a matter of convention, as long as we assume that this fraction is the same
for all particles. Such particles are referred to as dyons. By varying the mixing angle we can change the fraction of magnetic monopoles at will without
changing the laws of electrodynamics. For = 0 we recover the usual Maxwell
electrodynamics as we know it.
T HE
and hence F
(1.62)
3,
(1.63)
is conserved. I.e.,
1.4
19
E LECTROMAGNETIC D UALITY
E 2 c2 B2 = Const
E B = Const
(1.64a)
(1.64b)
(1.65)
= E E c2 B B + ic(E B) + ic(B E)
(1.66)
= 0 + 0 + ic(E B) ic(E B) = 0
= E E + c2 B B ic(E B) + ic(B E)
= 0 + 0 ic(E B) ic(E B) = 2ic(E B)
(1.67)
D UALITY EXPRESSED IN
EXAMPLE
1.3
Expressed in the complex field vector, introduced in Example 1.3 on the facing page,
the duality transformation Equations (1.52) on page 16 become
?
E XAMPLE
1.4
(1.68)
?F 2 = ei F ei F = |F|2
(1.69)
while
?
F ?F = e2i F F
(1.70)
Furthermore, assuming that = (t, x), we see that the spatial and temporal differenti-
20
C LASSICAL E LECTRODYNAMICS
ation of ?F leads to
?F
F
= i ei F + ei
t
t
t
?F ?F = iei F + ei F
t ?F
(1.71a)
(1.71b)
?F ?F = iei F + ei F
?F
(1.71c)
?F
OF EXAMPLE
1.4
Bibliography
[1] R. B ECKER, Electromagnetic Fields and Interactions, Dover Publications, Inc.,
New York, NY, 1982, ISBN 0-486-64290-9.
[2] W. G REINER, Classical Electrodynamics, Springer-Verlag, New York, Berlin,
Heidelberg, 1996, ISBN 0-387-94799-X.
[3] E. H ALLN, Electromagnetic Theory, Chapman & Hall, Ltd., London, 1962.
[4] J. D. JACKSON, Classical Electrodynamics, third ed., Wiley & Sons, Inc.,
New York, NY . . . , 1999, ISBN 0-471-30932-X.
[5] L. D. L ANDAU , AND E. M. L IFSHITZ, The Classical Theory of Fields,
fourth revised English ed., vol. 2 of Course of Theoretical Physics, Pergamon
Press, Ltd., Oxford . . . , 1975, ISBN 0-08-025072-6.
[6] J. C. M AXWELL, A Treatise on Electricity and Magnetism, third ed., vol. 1,
Dover Publications, Inc., New York, NY, 1954, ISBN 0-486-60636-8.
[7] D. B. M ELROSE , AND R. C. M C P HEDRAN, Electromagnetic Processes in Dispersive Media, Cambridge University Press, Cambridge . . . , 1991, ISBN 0-52141025-8.
[8] W. K. H. PANOFSKY, AND M. P HILLIPS, Classical Electricity and Magnetism,
second ed., Addison-Wesley Publishing Company, Inc., Reading, MA . . . , 1962,
ISBN 0-201-05702-6.
[9] J. S CHWINGER, A magnetic model of matter, Science 165 (1969).
[10] J. S CHWINGER , L. L. D E R AAD , J R ., K. A. M ILTON , AND W. T SAIYANG ,
Classical Electrodynamics, Perseus Books, Reading, MA, 1965, ISBN 0-73820056-5.
1.4
21
B IBLIOGRAPHY
22
C LASSICAL E LECTRODYNAMICS
2
Electromagnetic
Waves
In this chapter we shall investigate the dynamical properties of the electromagnetic field by deriving an set of equations which are alternatives to the Maxwell
equations. It turns out that these alternative equations are wave equations, indicating that electromagnetic waves are natural and common manifestations of
electrodynamics.
Maxwells microscopic equations (1.43) on page 14, which are usually
written in the following form
(t, x)
0
B
E =
t
B = 0
E =
B = 0 j(t, x) + 0 0
(2.1a)
(2.1b)
(2.1c)
E
t
(2.1d)
can be viewed as an axiomatic basis for classical electrodynamics. In particular, these equations are well suited for calculating the electric and magnetic
fields E and B from given, prescribed charge distributions (t, x) and current
distributions j(t, x) of arbitrary time- and space-dependent form.
However, as is well known from the theory of differential equations, these
four first order, coupled partial differential vector equations can be rewritten as
two un-coupled, second order partial equations, one for E and one for B. We
shall derive the second order equation for E, which, as we shall see is a homogeneous wave equation, and then discuss the implications of this equation. We
shall also show how the B field can be easily calculated from the solution of
the E equation.
23
24
E LECTROMAGNETIC WAVES
( E) = ( B) = 0
j + 0 E
t
t
t
(2.2)
(2.3)
(2.4)
and since EEMF = 0, Ohms law, Equation (1.26) on page 10, yields
j = E
(2.5)
E + 0 E = 0
E 0
t
t
2
(2.6)
1 2
E 2 2E = 0
t
c t
(2.7)
( E) = 0 E + 0 0 ( E) (2.8)
t
t
2.1
T HE
25
WAVE EQUATION
which, with the use of Equation (F.67) on page 155 and Equation (2.1c) on
page 23 can be rewritten
( B) 2 B = 0
2
B
0 0 2 B
t
t
(2.9)
Using the fact that, according to (2.1c), B = 0 for any medium and rearranging, we can rewrite this equation as
1 2
B 2 2B = 0
t
c t
2 B 0
(2.10)
This is the wave equation for the magnetic field. We notice that it is of exactly
the same form as the wave equation for the electric field, Equation (2.7) on the
preceding page.
(2.11)
and insert this into Equation (2.7) on the facing page. This yields
1 2
E0 (x)eit 2 2 E0 (x)eit
t
c t
1
= 2 E 0 (i)E0 (x)eit 2 (i)2 E0 (x)eit
c
1
2
2
= E 0 (i)E 2 (i) E = 0
c
2 E 0
2
1+i
E = 2 E + 2
0
c
1+
i
E=0
(2.12)
(2.13)
2
E=0
c2
(2.14)
26
E LECTROMAGNETIC WAVES
(2.15)
(2.16)
k=
c
1
=
=
=
0 0 ck k
0
= R0
0 k
(2.17)
where in the last step we introduced the characteristic impedance for vacuum
R0 =
0
376.7
0
(2.18)
(2.19)
E
n
B
n
B
n
=0
=
(2.20a)
B
t
(2.20b)
=0
= 0 j(t, x) + 0 0
(2.20c)
E
E
= 0 E + 0 0
t
t
(2.20d)
2.2
P LANE
27
WAVES
0 = n
n
= n
0 + 0 0
E
t
(2.21)
which simplifies to the first-order ordinary differential equation for the normal
component En of the electric field
dEn
+ En = 0
dt
0
(2.22)
(2.23)
This, together with (2.20a), shows that the longitudinal component of E, i.e.,
the component which is perpendicular to the plane surface is independent of
and has a time dependence which exhibits an exponential decay, with a decrement given by the relaxation time in the medium.
Scalar multiplying (2.20b) by n,
we similarly find that
0 = n
n
E
B
= n
t
(2.24)
or
n
B
=0
t
(2.25)
From this, and (2.20c), we conclude that the only longitudinal component of
B must be constant in both time and space. In other words, the only non-static
solution must consist of transverse components.
=0
0
2
t c2 t2
(2.26)
=0
2 c2 t2
(2.27)
28
E LECTROMAGNETIC WAVES
As is well known, each component of this equation has a solution which can
be written
Ei = f ( ct) + g( + ct),
i = 1, 2, 3
(2.28)
(2.29)
n = k
c
c
(2.30)
(2.31)
Let us consider the minus sign in the exponent in Equation (2.29) above.
This corresponds to a wave which propagates in the direction of increasing .
Inserting this solution into Equation (2.20b) on page 26, gives
n
E
= iB = ik n
E
(2.32)
k
1
1
n
E = k E = k E = 0 0 n
E
(2.33)
Hence, to each transverse component of E, there exists an associated magnetic field given by Equation (2.33) above. If E and/or B has a direction in
space which is constant in time, we have a plane polarised wave (or linearly
polarised wave).
2.2
P LANE
29
WAVES
E
+
i
E
=
+ K2E = 0
0
0
0
2
2
where
2
K = 0 0
2
1+i
= 2
0
c
(2.34)
1+i
= k2 1 + i
0
0
(2.35)
where, in the last step, Equation (2.16) on page 26 was used to introduce the
wave number k. Taking the square root of this expression, we obtain
K=k
1+i
= + i
0
(2.36)
1+i
= (2 2 ) + 2i
0
(2.37)
or
2 = 2 k2
(2.38)
(2.39)
k2
20
Squaring the latter and combining with the former, one obtains the second
order algebraic equation (in 2 )
2 (2 k2 ) =
k 4 2
420 2
(2.40)
=k
=k
1+
+1
(2.41a)
2
1+
(2.41b)
30
E LECTROMAGNETIC WAVES
(2.42)
With the aid of Equation (2.33) on page 28 we can calculate the associated
magnetic field, and find that it is given by
B=
1
1
1
K k E = ( k E)( + i) = ( k E) |A| ei
(2.43)
1
2
K = k 1+i
0
= 0 0 (1 + i)
0
=k i
1i
0
= (1 + i)
20
1
2
k(1 + i)
20
(2.44)
0
2
From this analysis we conclude that when the wave impinges perpendicularly upon the medium, the fields are given, inside this medium, by
E = E0 exp
0
B0 = (1 + i)
0
exp i
2
0
t
2
0
( n
E0 )
2
(2.45a)
(2.45b)
2
0
(2.46)
2.3
31
B IBLIOGRAPHY
always real valued. I.e., Ephysical = Re {Emathematical }. It is particularly important to remember this when one works with products of physical quantities. For instance, if we have two physical vectors F and G which both are
time-harmonic, i.e., can be represented by Fourier components proportional to
exp{it}, then we must make the following interpretation
F(t, x) G(t, x) = Re {F} Re {G} = Re F0 (x) eit ! Re G0 (x) eit ! (2.47)
Furthermore, letting denotes complex conjugate, we can express the real part
of the complex vector F as
1
Re {F} = Re F0 (x) eit ! = [F0 (x) eit + F0 (x) eit ]
2
(2.48)
(2.50)
Bibliography
[1] J. D. JACKSON, Classical Electrodynamics, third ed., Wiley & Sons, Inc.,
New York, NY . . . , 1999, ISBN 0-471-30932-X.
32
E LECTROMAGNETIC WAVES
3
Electromagnetic
Potentials
Instead of expressing the laws of electrodynamics in terms of electric and magnetic fields, it turns out that it is often more convenient to express the theory in
terms of potentials. In this Chapter we will introduce and study the properties
of such potentials.
(3.1)
Taking the divergence of this and using Equation (1.7) on page 4, we obtain
Poissons equation
2 stat (x) = Estat (x) =
(x)
0
(3.2)
1
40
(x0 ) 3 0
d x +
|x x0 |
(3.3)
where the integration is taken over all source points x0 at which the charge
density (x0 ) is non-zero and is an arbitrary quantity which has a vanishing
gradient. An example of such a quantity is a scalar constant. The scalar function stat (x) in Equation (3.3) above is called the electrostatic scalar potential.
33
34
E LECTROMAGNETIC P OTENTIALS
(3.4)
Bstat (x) =
(3.5)
0
4
j(x0 ) 3 0
d x + a(x)
|x x0 |
(3.6)
where a(x) is an arbitrary vector field whose curl vanishes. From Equation (M.78) on page 175 we know that such a vector can always be written
as the gradient of a scalar field.
3.3
T HE
35
(3.7)
Inserting this expression into the other homogeneous Maxwell equation, Equation (1.30) on page 11, we obtain
E(t, x) +
A(t, x) = 0
t
(3.8)
(3.9)
A(t, x) = (t, x)
(3.10)
t
This means that in electrodynamics, E(t, x) can be calculated from the formula
E(t, x) +
A(t, x)
(3.11)
t
and B(t, x) from Equation (3.7) above. Hence, it is a matter of taste whether we
want to express the laws of electrodynamics in terms of the potentials (t, x)
and A(t, x), or in terms of the fields E(t, x) and B(t, x). However, there exists an
important difference between the two approaches: in classical electrodynamics
the only directly observable quantities are the fields themselves (and quantities
derived from them) and not the potentials. On the other hand, the treatment
becomes significantly simpler if we use the potentials in our calculations and
then, at the final stage, use Equation (3.7) and Equation (3.11) above to calculate the fields or physical quantities expressed in the fields.
Inserting (3.11) and (3.7) into Maxwells equations (1.43) on page 14 we
obtain, after some simple algebra and the use of Equation (1.9) on page 5, the
general inhomogeneous wave equations
E(t, x) = (t, x)
1 2
1
(t, x)
A+ 2
+
2 =
c2 t2
0
t
c t
1 2
1
2
A A = 0 j(t, x) A + 2
2
2
c t
c t
(3.12a)
(3.12b)
36
E LECTROMAGNETIC P OTENTIALS
1
=0
c2 t
(3.13)
which is called the Lorentz gauge condition, because this choice simplifies the
system of coupled equations (3.12) on the previous page into the following set
of uncoupled partial differential equations which we call the Lorentz inhomogeneous wave equations:
1 2
(t, x)
2 =
c2 t2
0
2
def 1
2
A 2 2 A 2 A = 0 j(t, x)
c t
def
(3.14a)
(3.14b)
3.3
T HE
37
up to terms of vanishing gradients and curls, to pick suitable forms for the
potentials and still get the same physical result. In fact, the way the electromagnetic scalar potential (t, x) and the vector potential A(t, x) are related to
the physically observables gives leeway for similar manipulation of them
also in electrodynamics. If we transform (t, x) and A(t, x) simultaneously into
new ones 0 (t, x) and A0 (t, x) according to the mapping scheme
(t, x)
t
A(t, x) 7 A0 (t, x) = A(t, x) (t, x)
(t, x) 7 0 (t, x) = (t, x) +
(3.15a)
(3.15b)
E0 = 0 A0 = A + = A (3.16a)
t
t
t
t
t
0
0
B = A = A = A
(3.16b)
where, once again Equation (M.78) on page 175 was used. Comparing these
expressions with (3.11) and (3.7) we see that the fields are unaffected by the
gauge transformation (3.15). A transformation of the potentials and A which
leaves the fields, and hence Maxwells equations, invariant is called a gauge
transformation. A physical law which does not change under a gauge transformation is said to be gauge invariant. By definition, the fields themselves
are, of course, gauge invariant.
The potentials (t, x) and A(t, x) calculated from (3.12) on page 35, with
an arbitrary choice of A, can be further gauge transformed according to
(3.15) above. If, in particular, we choose A according to the Lorentz condition, Equation (3.13) on the facing page, and apply the gauge transformation
(3.15) on the resulting Lorentz equations (3.14) on the preceding page, these
equations will be transformed into
1 2
(t, x)
1 2
2
+
2 =
2
2
2
2
c t
t c t
0
2
2
1
1
A 2 A 2 2 2 = 0 j(t, x)
2
2
c t
c t
(3.17a)
(3.17b)
We notice that if we require that the gauge function (t, x) itself be restricted
to fulfil the wave equation
1 2
2 = 0
c2 t2
(3.18)
38
E LECTROMAGNETIC P OTENTIALS
these transformed Lorentz equations will keep their original form. The set of
potentials which have been gauge transformed according to Equation (3.15)
on the preceding page with a gauge function (t, x) which is restricted to fulfil
Equation (3.18) on the previous page, i.e., those gauge transformed potentials
for which the Lorentz equations (3.14) are invariant, comprises the Lorentz
gauge.
Other useful gauges are
The radiation gauge, also known as the transverse gauge, defined by
A = 0.
The Coulomb gauge, defined by = 0, A = 0.
The temporal gauge, also known as the Hamilton gauge, defined by
= 0.
The axial gauge, defined by A3 = 0.
The process of choosing a particular gauge condition is referred to as gauge
fixing.
(t, x) = f (t, x)
(3.19)
F 1 [ f (x)] f (t, x) =
def
f (x) eit d
(3.20a)
f (t, x) eit dt
(3.20b)
1
2
3.3
T HE
39
exists, and that the same is true for the generic potential component:
(t, x) =
(x) =
1
2
(x) eit d
(t, x) eit dt
(3.21a)
(3.21b)
Inserting the Fourier representations (3.20a) and (3.21a) into Equation (3.19)
on the preceding page, and using the vacuum dispersion relation for electromagnetic waves
= ck
(3.22)
(3.23)
(3.24)
and the solution of Equation (3.23) above which corresponds to the frequency
is given by the superposition
=
(3.25)
(3.26)
40
E LECTROMAGNETIC P OTENTIALS
Away from r = |x x0 | = 0, i.e., away from the source point x0 , this equation
takes the form
d2
(rG) + k2 (rG) = 0
dr2
with the well-known general solution
G = C+
(3.27)
ik|xx |
eikr
eikr
eik|xx |
e
+
C
+ C
C+
r
r
|x x0 |
|x x0 |
(3.28)
"
C +C
1
d3x0 + k2 C + + C #
"
|x x0 |
0
1
d3x0
|x x0 |
(3.30)
3 0
( x x ) d x
In virtue of the fact that the volume element d3x0 in spherical polar coordinates
is proportional to |x x0 |2 , the second integral vanishes when |x x0 | 0. Furthermore, from Equation (M.73) on page 174, we find that the integrand in the
first integral can be written as 4(|x x0 |) and, hence, that
1
(3.31)
4
Insertion of the general solution Equation (3.28) into Equation (3.25) on
the previous page gives
C+ + C =
ik|xx |
eik|xx | 3 0
0 e
d
x
+
C
d3x0
(3.32)
f (x )
f
(x
)
(x) = C
|x x0 |
|x x0 |
The Fourier transform to ordinary t domain of this is obtained by inserting the
above expression for (x) into Equation (3.21a) on the preceding page:
0
|
exp & i t k|xx
('
0
(t, x) = C
f (x )
%
+ C
%
f (x0 )
|x x0 |
d d3x0
|
exp & i t + k|xx
)'
(3.33)
|x x0 |
d d3x0
3.3
41
B IBLIOGRAPHY
0
0
tret
= tret
(t, x x0 ) = t
k |x x0 |
|x x0 |
= t
c
0
x0 |
k
x
|
|x
|x
0
0
= t+
tadv
= tadv
(t, x x0 ) = t +
(3.34a)
(3.34b)
0 , x0 )
f (tret
d3x0 + C
|x x0 |
0 , x0 )
f (tadv
d3x0
|x x0 |
(3.35)
This is a solution to the generic inhomogeneous wave equation for the potential
components Equation (3.19) on page 38. We note that the solution at time t at
the field point x is dependent on the behaviour at other times t 0 of the source
at x0 and that both retarded and advanced t 0 are mathematically acceptable
solutions. However, if we assume that causality requires that the potential at
0 , x0 ), we must in
(t, x) is set up by the source at an earlier time, i.e., at (tret
(t, x) =
(3.36a)
(3.36b)
Since these retarded potentials were obtained as solutions to the Lorentz equations (3.14) on page 36 they are valid in the Lorentz gauge but may be gauge
transformed according to the scheme described in Subsection 3.3.1 on page 36.
As they stand, we shall use them frequently in the following.
Bibliography
[1] L. D. FADEEV, AND A. A. S LAVNOV, Gauge Fields: Introduction to Quantum
Theory, No. 50 in Frontiers in Physics: A Lecture Note and Reprint Series. Ben-
42
E LECTROMAGNETIC P OTENTIALS
4
The
Electromagnetic
Fields
While, in principle, the electric and magnetic fields can be calculated from the
Maxwell equations in Chapter 1, or even from the wave equations in Chapter 2,
it is often physically more lucid to calculate them from the electromagnetic
potentials derived in Chapter 3. In this Chapter we will derive the electric and
magnetic fields from the potentials.
We recall that in order to find the solution (3.35) for the generic inhomogeneous wave equation (3.19) on page 38 we presupposed the existence of a
Fourier transform pair (3.20a) on page 38 for the generic source term
(t, x) =
(x) =
1
2
(x) eit d
(4.1a)
(t, x) eit dt
(4.1b)
That such transform pairs exists is true for most physical variables which are
not strictly monotonically increasing and decreasing with time. For charges
and currents varying in time we can therefore, without loss of generality, work
with individual Fourier components. Strictly speaking, the existence of a single
Fourier component assumes a monochromatic source (i.e., a source containing
only one single frequency component), which in turn requires that the electric
and magnetic fields exist for infinitely long times. However, by taking the
proper limits, we can still use this approach even for sources and fields of
finite duration.
This is the method we shall utilise in this Chapter in order to derive the
electric and magnetic fields in vacuum from arbitrary given charge densities
(t, x) and current densities j(t, x), defined by the temporal Fourier transform
43
44
E LECTROMAGNETIC P OTENTIALS
pairs
(t, x) =
(x) eit d
1
2
(x) =
(4.2a)
(t, x) eit dt
(4.2b)
and
j(t, x) =
j (x) =
j (x) eit d
1
2
(4.3a)
j(t, x) eit dt
(4.3b)
under the assumption that only retarded potentials produce physically acceptable solutions.1
The temporal Fourier transform pair for the retarded vector potential can
then be written
(t, x) =
(x) eit d
1
(x) =
2
1
(t, x) e dt =
4
0
it
(4.4a)
0
eik|xx | 3 0
(x )
dx
|x x0 |
0
(4.4b)
where in the last step, we made use of the explicit expression for the temporal
Fourier transform of the generic potential component (x), Equation (3.32)
on page 40. Similarly, the following Fourier transform pair for the vector potential must exist:
A(t, x) =
1
A (x) =
2
A (x) eit d
(4.5a)
0
eik|xx | 3 0
0
j (x0 )
dx
A(t, x) e dt =
4
|x x0 |
it
(4.5b)
(4.6)
in order that all physical quantities be real. Similar transform pairs and requirements of real-valuedness exist for the fields themselves.
1 In
fact, John A. Wheeler and Richard P. Feynman derived in 1945 a fully self-consistent
electrodynamics using both the retarded and the advanced potentials [6]; See also [1].
4.1
T HE
45
MAGNETIC FIELD
(4.7a)
i0 t
(4.7b)
(t, x) = 0 (x)ei0 t
(4.7c)
i0 t
(4.7d)
j(t, x) = j0 (x)e
A(t, x) = A0 (x)e
(4.8)
(4.9)
which provides a relation between (the Fourier transforms of) the vector and
scalar potentials.
Using the Fourier transformed version of Equation (4.8) and Equation (4.5b)
on the preceding page, we obtain
0
B (x) = A (x) =
4
eik|xx | 3 0
j (x )
dx
|x x0 |
V
0
(4.10)
46
T HE E LECTROMAGNETIC F IELDS
eik|xx |
d3x0
0 |
x
|x
V
x x0
0
0
eik|xx | d3x0
j (x )
3
0
V
|x 0x |
x x ik|xx0 |
1
3 0
+ j (x0 ) ik
e
0
|x x |
|x x0 | d x
V
j (x )
0
B (x) =
4
0
=
4
(4.11)
j (x0 )eik|xx | (x x0 ) 3 0
dx
|x x0 |3
V
0
(ik)j(x0 )eik|xx | (x x0 ) 3 0
dx
+
|x x0 |2
V
0
4
From this expression for the magnetic field in the frequency () domain,
we obtain the magnetic field in the temporal (t) domain by taking the inverse
Fourier transform (using the identity ik = i/c):
B(t, x) =
0
=
4
B (x) eit d
d3x0
|x x0 |3
0
(i)j (x0 )ei(k|xx |t) d (x x0 )
1
c V
|x x0 |2
0 , x0 ) (x x0 )
0
j(tret
=
d3x0
3
0
4
|x+-, x |
* V
.
+
0
+
* 4c
d3x0
(4.12)
Induction field
0 , x0 ) (x x0 )
j(tret
d3x0
02
|x+/
, x|
Radiation field
where
def
0
j(tret
, x0 )
j
t
(4.13)
0
t=tret
The first term, the induction field, dominates near the current source but falls
off rapidly with distance from it, is the electrodynamic version of the BiotSavart law in electrostatics, formula (1.13) on page 6. The second term, the
radiation field or the far field, dominates at large distances and represents energy that is transported out to infinity. Note how the spatial derivatives ()
gave rise to a time derivative ()!
4.2
T HE
47
ELECTRIC FIELD
1
eik|xx | 3 0 i0
dx +
(x0 )
40 V
4
|x x0 |
0|
0
ik|xx
0
(x )e
(x x ) 3 0
1
dx
=
3
0
40 V
|x x |
ik
j (x0 )
eik|xx | 3 0
dx
|x x0 |
(4.14)
dx
c |x x0 |
|x x0 |
(4.15)
Doing so in the last term of Equation (4.14), and also using the fact that k =
/c, we can rewrite this Equation as
1
40
E (x) =
1
c *
(x0 )eik|xx | (x x0 ) 3 0
dx
|x x0 |3
0
0
[ j (x0 )](x x0 )
eik|xx | 3 0
0
ikj
(x
)
|x x0 | d x.
|x x0 |
+-,
(4.17)
I
The last integral can be further rewritten in the following way:
I =
=
But, since
0
xm
V
[0 j (x0 )](x x0 )
eik|xx | 3 0
0
ikj
(x
)
|x x0 |
|x x0 | d x
0
eik|xx | 3 0
jm xl xl0
0
ik jl (x ) x l
0 |x x0 |
xm
|x x0 | d x
xl xl0 ik|xx0 |
jm
e
=
|x x0 |2
(4.18)
jm xl xl0 ik|xx0 |
e
0
xm
|x x0 |2
xl xl0 ik|xx0 |
+ jm 0
e
xm |x x0 |2
(4.19)
48
T HE E LECTROMAGNETIC F IELDS
we can rewrite I as
I =
0
ik|xx |
jm x0 xl x0 l2 x l eik|xx0 | + ikj e|x x0 | d3x0
m |x x0 |
xl xl
0
jm
x l eik|xx | d3x0
2
0
0
xm
|x x |
(4.20)
I =
ik
0
eik|xx |
2
j
j (x x0 ) (x x0 )eik|xx | d3x0
+
2
4
0
0
|x x |
|x x |
V0
j (x x0 ) (x x0 )
|x x0 |3
ik|xx0 |
eik|xx |
+ j
d3x0
|x x0 | 1
(4.21)
Using the triple product bac-cab formula (F.54) on page 155 backwards, and
inserting the resulting expression for I into Equation (4.17) on the preceding
page, we arrive at the following final expression for the Fourier transform of
the total E-field:
0
1
eik|xx | 3 0 i0
E (x) =
dx +
(x0 )
40 V
4
|x x0 |
0|
0
ik|xx
0
1
(x )e
(x x ) 3 0
=
dx
3
40 V
|x x0 |
0
dx
c V
|x x0 |3
+
1
c
eik|xx | 3 0
dx
j (x )
|x x0 |
V
0
(4.22)
Taking the inverse Fourier transform of Equation (4.22) above, once again
using the vacuum relation = kc, we find, at last, the expression in time do-
4.2
49
B IBLIOGRAPHY
E (x) eit d
1
4
* 0
0 , x0 )(x x0 )
(tret
d3x0
3
0
|x+-, x |
.
1
* 40 c
0 , x0 ) (x x0 )](x x0 )
[j(tret
d3x0
0 |4
x
|x
+/,
.
(4.23)
Intermediate field
1
* 40 c
0 , x0 ) (x x0 )] (x x0 )
[j(tret
d3x0
+-|x, x0 |4
.
Intermediate field
1
* 40 c2
0 , x0 ) (x x0 )] (x x0 )
[j(tret
d3x0
+/|x, x0 |3
.
Radiation field
Here, the first term represents the retarded Coulomb field and the last term
represents the radiation field which carries energy over very large distances.
The other two terms represent an intermediate field which contributes only in
the near zone and must be taken into account there.
With this we have achieved our goal of finding closed-form analytic expressions for the electric and magnetic fields when the sources of the fields
are completely arbitrary, prescribed distributions of charges and currents. The
only assumption made is that the advanced potentials have been discarded;
recall the discussion following Equation (3.35) on page 41 in Chapter 3.
Bibliography
[1] F. H OYLE , S IR , AND J. V. NARLIKAR, Lectures on Cosmology and Action at a
Distance Electrodynamics, World Scientific Publishing Co. Pte. Ltd, Singapore,
New Jersey, London and Hong Kong, 1996, ISBN 9810-02-2573-3(pbk).
[2] J. D. JACKSON, Classical Electrodynamics, third ed., Wiley & Sons, Inc.,
New York, NY . . . , 1999, ISBN 0-471-30932-X.
[3] L. D. L ANDAU , AND E. M. L IFSHITZ, The Classical Theory of Fields, fourth revised English ed., vol. 2 of Course of Theoretical Physics, Pergamon Press, Ltd.,
Oxford . . . , 1975, ISBN 0-08-025072-6.
50
T HE E LECTROMAGNETIC F IELDS
5
Relativistic
Electrodynamics
propagation that equals the speed of light c = 1/ 0 0 and which can be considered as a constant of nature. To take this finite speed of propagation of
information into account, and to ensure that our laws of physics be independent of any specific coordinate frame, requires a treatment of electrodynamics
in a relativistically covariant (coordinate independent) form. This is the object
of the current chapter.
51
52
R ELATIVISTIC E LECTRODYNAMICS
vt
y
0
y0
v
P(t, x, y, z)
P(t0 , x0 , y0 , z0 )
O0
x0
z0
F IGURE 5.1: Two inertial systems and 0 in relative motion with velocity v along the x = x0 axis. At time t = t0 = 0 the origin O0 of 0 coincided
with the origin O of . At time t, the inertial system 0 has been translated
a distance vt along the x axis in . An event represented by P(t, x, y, z) in
is represented by P(t0 , x0 , y0 , z0 ) in 0 .
v
c
(5.1)
1
1 2
(5.2)
where v = |v|. In the following, we shall make frequent use of these shorthand
notations.
As shown by Einstein, the two postulates of special relativity require that
the spatial coordinates and times as measured by an observer in and 0 ,
5.1
T HE
53
(5.3a)
(5.3b)
(5.3c)
(5.3d)
x = (x vt)
y =y
z =z
Taking the difference between the square of (5.3a) and the square of (5.3b) we
find that
c2 t02 x02 = 2 c2 t2 2xct + x2 2 x2 + 2xvt v2 t2 #
"
c t
2 2
v2
1 2
c
= c2 t 2 x 2
v2
v2
1 2 x2 1 2
c
c
(5.4)
From Equations (5.3) we see that the y and z coordinates are unaffected by
the translational motion of the inertial system 0 along the x axis of system
. Using this fact, we find that we can generalise the result in Equation (5.4)
above to
c2 t2 x2 y2 z2 = c2 t02 x02 y02 z02
(5.5)
which means that if a light wave is transmitted from the coinciding origins O
and O0 at time t = t0 = 0 it will arrive at an observer at (x, y, z) at time t in
and an observer at (x0 , y0 , z0 ) at time t0 in 0 in such a way that both observers
conclude that the speed (spatial distance divided by time) of light in vacuum
is c. Hence, the speed of light in and 0 is the same. A linear coordinate transformation which has this property is called a (homogeneous) Lorentz
transformation.
(5.6)
We then interpret this quadruple x as (the component form of) a radius fourvector in a real, linear, four-dimensional vector space.
54
R ELATIVISTIC E LECTRODYNAMICS
Metric tensor
We want our space to be a Riemannian space, i.e., a space where a distance
and a scalar product are defined. We therefore need to define in this space a
metric tensor, also known as the fundamental tensor, which we shall denote
g and choose as (in matrix notation):
6-77
(g ) =
3445
1 0
0
0
0 1 0
0
0 0 1 0 8
0 0
0 1
(5.7)
i.e., with a main diagonal with sign sequence, or signature, {+, , , }.1
Radius four-vector in contravariant and covariant form
The radius four-vector x = (x0 , x1 , x2 , x3 ) = (ct, x), as defined in Equation (5.6)
on the previous page, is, by definition, the prototype of a contravariant vector
(or, more accurately, a vector in contravariant component form). The corresponding covariant vector x is obtained as (the upper index in x is summed
over and is therefore a dummy index and may be replaced by another dummy
index ):
x = g x
(5.8)
6-77
3445
x0
x1
x2 8
x3
6-77
3445
1 0
0
0
0 1 0
0
0 0 1 0 8
0 0
0 1
3445
6-77
x0
x1
x2 8
x3
6-77
3445
x0
x1
x2 8
x3
(5.9)
1 Without changing the physics, one can alternatively choose a signature {, +, +, +}. The latter has the advantage that the transition from 3D to 4D becomes smooth, while it will introduce
some annoying minus signs in the theory. In current physics literature, the signature {+, , , }
seems to be the most commonly used one.
5.1
T HE
55
(5.10)
i.e., the covariant radius four-vector x is obtained from the contravariant radius four-vector x simply by changing the sign of the last three components.
These components are referred to as the space components; the zeroth component is referred to as the time component.
Scalar product and norm
Taking the scalar product of x with itself, we get, by definition,
g x x = x x = (x0 , x1 , x2 , x3 ) (x0 , x1 , x2 , x3 ) = (ct, x) (ct, x)
= (ct, x, y, z) (ct, x, y, z) = c2 t2 x2 y2 z2
(5.11)
(5.12a)
g = g
g g =
g g =
g
g
(5.12b)
=
=
(5.12c)
(5.12d)
0 if 6=
1 if =
(5.13)
56
R ELATIVISTIC E LECTRODYNAMICS
(5.14)
where the metric tensor is as in Equation (5.7) on page 54. As we see, this
form is indefinite as expected for a non-Euclidian space. The square root of
this expression is the invariant line element
ds = c
1
1 2
c
dx 1
dt
dx 2
+
dt
dx 3
+
dt
;
dt
v2
1
(v x )2 + (vy )2 + (vz )2 dt = c 1 2 dt
2
c
c
c
1 2 dt = dt = c d
= c 1
=c
(5.15)
where we introduced
d = dt/
(5.16)
Since d measures the time when no spatial changes are present, it is called the
proper time.
Expressing Equation (5.5) on page 53 in terms of the differential interval
ds and comparing with Equation (5.14), we find that
ds2 = c2 dt2 dx2 dy2 dz2
(5.17)
(5.18)
(5.19)
(5.20)
5.1
T HE
57
is a light-like interval; we may also say that in this case we are on the light
cone. A vector which has a light-like interval is called a null vector. The
time-like, space-like or light-like aspects of an interval ds is invariant under a
Lorentz transformation.
Four-vector fields
Any quantity which relative to any coordinate system has a quadruple of real
numbers and transforms in the same way as the radius four-vector x does, is
called a four-vector. In analogy with the notation for the radius four-vector we
introduce the notation a = (a0 , a) for a general contravariant four-vector field
in 9 4 and find that the lowering of index rule, Equation (M.20) on page 163,
for such an arbitrary four-vector yields the dual covariant four-vector field
a (x ) = g a (x ) = (a0 (x ), a(x ))
(5.21)
The scalar product between this four-vector field and another one b (x ) is
g a (x )b (x ) = (a0 , a) (b0 , b) = a0 b0 a b
(5.22)
"
6-77
3445
0 0
0 0
0
0
1 08
0
0
0 1
(5.23)
the linear Lorentz transformation (5.3) on page 53, i.e., the coordinate transformation x x0 = x0 (x0 , x1 , x2 , x3 ), from one inertial system to another
inertial system 0 , can be written
x0 = x
(5.24)
(5.25)
58
R ELATIVISTIC E LECTRODYNAMICS
1 + 2
1 + 1 2
(5.26)
This means that the nonempty set of Lorentz transformations constitute a closed
algebraic structure with a binary operation which is associative. Furthermore,
one can show that this set possesses at least one identity element and at least
one inverse element. In other words, this set of Lorentz transformations constitute a mathematical group. However tempting, we shall not make any further
use of group theory.
(5.27a)
(5.27b)
(5.27c)
(5.27d)
X =x
X =x
X =x
dS = ids
(5.27e)
(5.28)
5.1
T HE
59
X0
X 00
x01
x1
w
x0 = ct
x00
x0 = x 1
P0
x01
ct
O = O0
P
x1 = x
F IGURE 5.3: Minkowski diagram depicting geometrically the transformation (5.31) from the unprimed system to the primed system. Here w denotes the world line for an event and the line x0 = x1 x = ct the world
line for a light ray in vacuum. Note that the event P is simultaneous with
all points on the x1 axis (t = 0), including the origin O while the event
P0 , which is also simultaneous with all points on the x0 axis, including
O0 = O, to an observer at rest in the primed system, is not simultaneous
with O in the unprimed system but occurs there at time |P P0 | /c.
60
R ELATIVISTIC E LECTRODYNAMICS
(5.29)
(5.30a)
x = x cos + X sin
(5.30b)
01
x = x cosh ct sinh
(5.31a)
(5.31b)
(5.32a)
cosh =
(5.32b)
tanh =
(5.32c)
It is therefore possible to envisage the Lorentz transformation as an ordinary rotation in the 4D Euclidean space < 4 This rotation i < 4 corrsponds to a
coordinate change in 9 4 as depicted in Figure 5.3 on the preceding page. Equation (5.26) on page 58 for successive Lorentz transformation then corresponds
to the tanh addition formula
tanh(1 + 2 ) =
tanh 1 + tanh 2
1 + tanh 1 tanh 2
(5.33)
The use of ict and < 4 , which leads to the interpretation of the Lorentz
transformation as an ordinary rotation, may, at best, be illustrative, but is
not very physical. Besides, if we leave the flat 9 4 space and enter the curved
space of general relativity, the ict trick will turn out to be an impasse. Let us
therefore immediately return to 9 4 where all components are real valued.
5.2
C OVARIANT
61
CLASSICAL MECHANICS
u =
dx
v
= 1, =
ds
c
53 ?
1
1
v2
c2
v
v2
c2
c 1
(5.34)
which, when multiplied with the scalar invariant m0 c2 yields the four-momentum
p = m0 c2
dx
= m0 c (c, v) =
ds
53 ?
m0
c2
v2
c2
m0 cv
1
v2
c2
(5.35)
(5.36)
where
m = m0 =
m0
(5.37)
1 vc2
i.e., that Lorentz covariance implies that the mass-like term in the ordinary 3D
linear momentum is not invariant.
The zeroth (time) component of the four-momentum p is given by
p0 = m0 c2 =
m0 c2
1
v2
c2
= mc2
(5.38)
(5.39)
(m0 c2 )2
2
1 cv2
(5.40)
v2
1 2 = (m0 c2 )2
c
Since this is an invariant, this equation holds in any inertial frame, particularly
in the frame where p = 0 where we thus have
E = m 0 c2
(5.41)
62
R ELATIVISTIC E LECTRODYNAMICS
dx
v
= ,
ds
c
(5.42)
where we introduced
= 0
(5.43)
= = =
1 2
2
c2 t2
(5.44)
Since it has the characteristics of a four-scalar, the dAlembert operator is invariant and, hence, the homogeneous wave equation is Lorentz covariant.
(5.45)
where is the scalar potential and A the vector potential, defined in Section 3.3
on page 34, we can write the inhomogeneous wave equations (Lorentz equations) Equation (3.14) on page 36 in the following compact (and covariant)
way:
A =
j
0
(5.46)
With the help of the above, we can formulate our electrodynamic equations
covariantly. For instance, the covariant form of the equation of continuity,
Equation (1.21) on page 9 is
j
j
=0
x
(5.47)
and the Lorentz gauge condition, Equation (3.13) on page 36, can be written
A
A
=0
x
(5.48)
5.3
C OVARIANT
63
CLASSICAL ELECTRODYNAMICS
c(x )
x
(5.49)
If only one dimension Lorentz contracts (for instance, due to relative motion along the x direction), a 3D spatial volume transforms according to
2
v2
1
dV = d3x = dV0 = dV0 1 2 = dV0 1 2
(5.50)
(5.51)
i.e., the charge in a given volume is conserved. We can therefore conclude that
the electron charge is a universal constant.
"
=
0
q0
1
,0
40 |x x0 |0
(5.52)
where |x x0 |0 is the usual distance from the source point to the field point,
evaluated in the rest system (signified by the index 0).
Let us introduce the relative radius four-vector between the source point
and the field point:
R = x x0 = (c(t t0 ), x x0 )
(5.53)
2
(5.54)
We know that in vacuum the signal (field) from the charge q0 at x0 propagates to x with the speed of light c so that
x x0 = c(t t0 )
(5.55)
64
R ELATIVISTIC E LECTRODYNAMICS
Inserting this into Equation (5.54) on the preceding page, we see that
R R = 0
(5.56)
R = ( x x 0 , x x 0 )
(5.57)
Now we want to find the correspondence to the rest system solution, Equation (5.52) on the preceding page, in an arbitrary inertial system. We note from
Equation (5.34) on page 61 that in the rest system
"
u #
53 ?
and
1
1
v2
c2
v
1
v2
c2
= (1, 0)
(5.58)
(v=0)
(R )0 = ( x x0 , x x0 )0 = ( x x0 0 , (x x0 )0 )
(5.59)
"
= (u )0 (R )0
= (1, 0) ( x x0 0 , (x x0 )0 ) = x x0 0
(5.60)
q0 u
40 u R
(5.61)
0
v
0
c ( x x , (x x ))
(x x0 ) v
= x x0
c
u R = 1,
(5.62)
Introducing
s = x x0
(x x0 ) v
c
(5.63)
5.3
C OVARIANT
65
CLASSICAL ELECTRODYNAMICS
we can write
u R = s
and
u
=
u R
(5.64)
1 v
,
s cs
(5.65)
q0
40
1 v
,
= (, cA)
s cs
(5.66)
where in the last step the definition of the four-potential, Equation (5.45) on
page 62, was used. Writing the solution in the ordinary 3D-way, we conclude
that for a very localised charge volume, moving relative an observer with a
velocity v, the scalar and vector potentials are given by the expressions
q0
1
q0 1
=
0
0
40 s 40 |x x | (xxc )v
v
q0 v
q0
A(t, x) =
=
0
2
2
0
40 c s 40 c |x x | (xxc )v
(t, x) =
(5.67a)
(5.67b)
= (a2 b3 a3 b2 ) x 1
(5.68)
+ (a3 b1 a1 b3 ) x 2
+ (a1 b2 a2 b1 ) x 3
i, j 6= k
(5.69)
In other words, the pseudovector c = a b can be considered as an antisymmetric tensor of rank two!
66
R ELATIVISTIC E LECTRODYNAMICS
The same is true for the curl operator . For instance, the Maxwell equation
E =
B
t
(5.70)
(5.71)
We know from Chapter 3 that the fields can be derived from the electromagnetic potentials in the following way:
B = A
E =
(5.72a)
A
t
(5.72b)
= i A j j Ai
xi x j
Ai
Ei = i
= i t Ai
x
t
(5.73a)
Bi j =
(5.73b)
From this, we notice the clear difference between the axial vector (pseudovector)
B and the polar vector (ordinary vector) E.
Our goal is to express the electric and magnetic fields in a tensor form
where the components are functions of the covariant form of the four-potential,
Equation (5.45) on page 62:
A = (, cA)
(5.74)
Inspection of (5.74) and Equation (5.73) above makes it natural to define the
covariant four-tensor
F =
A A
= A A
x x
(5.75)
6-77
F = A A =
3445
0
Ex
Ey
Ez
E x
0
cBz cBy
Ey cBz
0
cB x 8
Ez cBy cB x
0
(5.76)
5.3
67
B IBLIOGRAPHY
6-77
F = A A =
3445
0 E x Ey Ez
Ex
0
cBz cBy
Ey cBz
0
cB x 8
Ez cBy cB x
0
(5.77)
B = 0 j + 0
E
E
= 0 v + 0
t
t
(5.78)
(5.79)
correspond to
F
j
=
x
0
(5.80)
B
t
B = 0
(5.81)
(5.82)
correspond to
F F F
+ +
=0
x
x
x
(5.83)
Hence, Equation (5.80) and Equation (5.83) above constitute Maxwells equations in four-dimensional formalism.
Bibliography
[1] J. A HARONI, The Special Theory of Relativity, second, revised ed., Dover Publications, Inc., New York, 1985, ISBN 0-486-64870-2.
[2] A. O. BARUT, Dynamics and Classical Theory of Fields and Particles, Dover
Publications, Inc., New York, NY, 1980, ISBN 0-486-64038-8.
68
R ELATIVISTIC E LECTRODYNAMICS
6
Interactions of
Fields and
Particles
In this Chapter we study the interaction between electric and magnetic fields
and electrically charged particles. The analysis is based on Lagrangian and
Hamiltonian methods, is fully covariant, and yields results which are relativistically correct.
x1
x0
L(4) (x , u ) ds = 0
(6.1)
where ds is the invariant line element given by Equation (5.15) on page 56,
and the endpoints are fixed.
69
70
I NTERACTIONS
OF
F IELDS
AND
PARTICLES
free
L(4)
=
(6.2)
For an interaction with the electromagnetic field we can introduce the interaction with the help of the four-potential given by Equation (5.74) on page 66 in
the following way
L(4) =
m0 c2
u u + qu A (x )
2
(6.3)
x1
L(4) (x , u ) ds =
x0
x1
=
=
x0
x1
x0
x1
x0
m0 c2
u u + qu A ds
2
A
m0 c2
u u # u + q A u + u x ds
2 u "
x
mo c u u
2
+ q A u + u
(6.4)
A
x
ds = 0
x
dx
ds
(6.5)
which means that we can write the variation of u as a total derivative with
respect to s :
u =
dx
d
=
x #
ds
ds "
(6.6)
6.1
C HARGED PARTICLES
IN AN
71
E LECTROMAGNETIC F IELD
Inserting this into the first two terms in the last integral in Equation (6.4) on
the facing page, we obtain
x1
L(4) (x , u ) ds
x0
A
d
d
mo c u
x # + qA
x # + qu x ds
ds "
ds "
x
x1
(6.7)
x0
Partial integration in the two first terms in the right hand member of (6.7)
gives
x1
L(4) (x , u ) ds
x0
x1
dA
A
m0 c
x q
x + qu x ds
ds
ds
x
2 du
x0
(6.8)
where the integrated parts do not contribute since the variations at the endpoints vanish. A change of irrelevant summation index from to in the first
two terms of the right hand member of (6.8) yields, after moving the common
factor x outside the partenthesis, the following expression:
x1
L(4) (x , u ) ds
x0
x1
x0
A
dA
m0 c
q
+ qu x ds
ds
ds
x
2 du
(6.9)
(6.10)
By inserting this expression (6.10) into the second term in right-hand member
of Equation (6.9) above, and noting the common factor qu of the resulting
term and the last term, we obtain the final variational principle expression
x1
L(4) (x , u )ds
x0
x1
x0
m0 c
2 du
ds
+ qu
A A
x ds
x x
(6.11)
72
I NTERACTIONS
OF
F IELDS
AND
PARTICLES
Since, according to the variational principle, this expression shall vanish and
x is arbitrary along the world-line between the fixed end points x0 and x1 , the
expression inside in the integrand in the right hand member of (6.11) must
vanish. In other words,
we have found an equation of motion for a charged
particle in a prescribed electromagnetic field:
m0 c
2 du
ds
= qu
A A
= qu A A #
"
x x
(6.12)
With the help of Equation (5.75) on page 66 we can express this equation in
terms of the electromagnetic field tensor in the following way:
m0 c2
du
= qu F
ds
(6.13)
(6.14a)
(6.14b)
L(4)
u
(6.15)
(6.16)
6.1
C HARGED PARTICLES
IN AN
73
E LECTROMAGNETIC F IELD
With the help of these, the radius four-vector x , considered as the generalised
four-coordinate, and the invariant line element ds, defined in Equation (5.15)
on page 56, we introduce the following eight partial differential equations:
H(4) dx
=
(6.17a)
p
ds
dp
H(4)
=
(6.17b)
x
ds
which form the four-dimensional Hamilton equations.
Our strategy now is to use Equation (6.15) on the facing page and Equations (6.17) above to derive an explicit algebraic expression for the canonically
conjugate momentum four-vector. According to Equation (5.39) on page 61, a
four-momentum has a zeroth (time) component which we can identify with the
total energy. Hence we require that the component p0 of the conjugate fourmomentum vector defined according to Equation (6.15) on the preceding page
be identical to the ordinary 3D Hamiltonian H and hence that this component
solves the Hamilton equations Equations (6.14) on the facing page. This later
consistency check is left as an exercise to the reader.
Using the definition of H(4) , Equation (6.16) on the preceding page, and
the expression for L(4) , Equation (6.3) on page 70, we obtain
m0 c2
u u qu A (x )
(6.18)
2
Furthermore, from the definition (6.15) of the conjugate four-momentum p ,
we see that
H(4) = p u L(4) = p u
L(4)
p =
=
u
u
= m0 c u + qA
m0 c2
u u + qu A (x )
2
(6.19)
m0 c2
u u qu A (x )
2
(6.20)
m0 c2
u u
=
2
Since the four-velocity scalar-multiplied by itself is u u = 1, we clearly
see from Equation (6.20) above that H(4) is indeed a scalar invariant, whose
value is simply
H(4) =
m0 c2
2
(6.21)
74
I NTERACTIONS
OF
F IELDS
AND
PARTICLES
1
p qA #
m0 c2 "
(6.22)
1
1
m0 c2
H(4) =
p qA #
p qA #
2
2
m0 c "
m0 c2 "
1
=
p qA # p qA #
"
2m0 c2 "
1
=
p p 2qA p + q2 A A #
2m0 c2 "
(6.23)
That this four-Hamiltonian yields the correct covariant equation of motion can
be seen by inserting it into the four-dimensional Hamiltons equations (6.17)
and using the relation (6.22):
q
A
H(4)
=
(p qA )
2
x
m0 c
x
q
A
=
m0 c2 u
m0 c2
x
(6.24)
A
= qu
x
du
A
dp
= mo c2
q u
=
ds
ds
x
where in the last step Equation (6.19) on the preceding page was used. Rearranging terms, and using Equation (5.77) on page 67, we obtain
m0 c
2 du
ds
= qu
A A
= qu F
x x
(6.25)
A = (, cA)
(6.26b)
p p = (p0 )2 c2 (p)2
A p = p c (p A)
A A = 2 c2 (A)2
(6.26a)
(6.26c)
(6.26d)
(6.26e)
6.1
C HARGED PARTICLES
IN AN
75
E LECTROMAGNETIC F IELD
(6.29)
(p qA)2 + m20 c2
p = q c
(6.30)
(p qA)2 + m20 c2
(6.31)
(6.32)
Using the explicit expressions (Equation (6.31)) and (Equation (6.32) above),
we obtain the explicit expression for the ordinary 3D Lagrange function
L = p v q c
(p qA)2 + m20 c2
(6.33)
m0 v
1
v2
c2
= mv
(6.34)
76
I NTERACTIONS
i1
m
k
a
i
m
k
a
OF
F IELDS
AND
PARTICLES
i+1
k
a
k
a
where the quantity mv is called the kinetic momentum, we can rewrite this
expression for the ordinary Lagrangian as follows:
L = qA v + mv q c m2 v2 m20 c2
= q + qA v m0 c
v2
1 2
c
(6.35)
What we have obtained is the correct expression for the Lagrangian describing
the motion of a charged particle in scalar and vector potentials associated with
prescribed electric and magnetic fields.
6.2
77
(6.36)
1 N
m2i k(i+1 i )2
2
i=1
(6.37)
L = a@
(6.38)
i=1
Here,
i+1 i
1 m 2
i ka
2 a
a
(6.39)
is the so called linear Lagrange density. If we now let N and, at the same
time, let the springs become infinitesimally short according to the following
scheme:
a dx
m
dm
=
a
dx
ka Y
i+1 i
a
x
(6.40a)
linear mass density
(6.40b)
Youngs modulus
(6.40c)
(6.40d)
78
I NTERACTIONS
F IELDS
OF
AND
PARTICLES
we obtain
L=
where
@
@
dx
(6.41)
, , , t = :
t x
2
t
x
;
(6.42)
Notice how we made a transition from a discrete description, in which the mass
points were identified by a discrete integer variable i = 1, 2, . . . , N, to a continuous description, where the infinitesimal mass points were instead identified by
a continuous real parameter x, namely their position along x .
A consequence of this transition is that the number of degrees of freedom
for the system went from the finite number N to infinity! Another consequence
is that @ has now become dependent also on the partial derivative with respect
to x of the field coordinate . But, as we shall see, the transition is well
worth the price because it allows us to treat all fields, be it classical scalar
or vectorial fields, or wave functions, spinors and other fields that appear in
quantum physics, on an equal footing.
Under the assumption of time independence and fixed endpoints, the variation principle (6.36) on the previous page yields:
L dt
=
=
A
ACBD
,
,
dx dt
t x
@
@
dx dt
t
x EF
t
x
(6.43)
=0
The last integral can be integrated by parts. This results in the expression
ACBD
53
53
8
EF
dx dt = 0
(6.44)
where the variation is arbitrary (and the endpoints fixed). This means that the
integrand itself must vanish. If we introduce the functional derivative
@
@
53
(6.45)
6.2
79
53
8
=0
(6.46)
2
2
2Y 2 =
t
x
2
2
=0
Y t2 x2
(6.47)
i.e., the one-dimensional wave equation for compression waves which propag
ate with phase speed v = Y/ along the linear structure.
A generalisation of the above 1D results to a three-dimensional continuum
is straightforward. For this 3D case we get the variational principle
L dt =
=
=
@
d3x dt
G
d4x
x
A BD
53
8
EF
(6.48)
4
d x
=0
where the variation is arbitrary and the endpoints are fixed. This means that
the integrand itself must vanish:
@
53
=0
(6.49)
xi
53
xi
(6.50)
80
I NTERACTIONS
53
F IELDS
AND
PARTICLES
OF
8
=0
(6.51)
(6.52)
, , i ; t = @
x
t
,
,
t xi
(6.53)
If, as usual, we differentiate this expression and identify terms, we obtain the
following Hamilton density equations
t
H
(6.54a)
(6.54b)
The Hamilton density functions are in many ways similar to the ordinary
Hamilton functions and lead to similar results.
The electromagnetic field
Above, when we described the mechanical field, we used a scalar field (t, x).
If we want to describe the electromagnetic field in terms of a Lagrange density
@ and Euler-Lagrange equations, it comes natural to express @ in terms of
the four-potential A (x ).
The entire system of particles and fields consists of a mechanical part, a
field part and an interaction part. We therefore assume that the total Lagrange
density @ tot for this system can be expressed as
tot
=@
mech
+@
inter
+@
field
(6.55)
where the mechanical part has to do with the particle motion (kinetic energy).
It is given by L(4) /V where L(4) is given by Equation (6.2) on page 70 and V is
6.2
81
the volume. Expressed in the rest mass density 0m , the mechanical Lagrange
density can be written
mech
1
= 0m c2 u u
2
(6.56)
The @ inter part which describes the interaction between the charged particles
and the external electromagnetic field. A convenient expression for this interaction Lagrange density is
inter
= j A
(6.57)
For the field part @ field we choose the difference between magnetic and
electric energy density (in analogy with the difference between kinetic and
potential energy in a mechanical field). Using the field tensor, we express this
field Lagrange density as
field
1
= 0 F F
4
(6.58)
tot
1
1
= 0m c2 u u + j A + 0 F F
2
4
(6.59)
(6.60)
IJJK
F = A A =
0
Ex
Ey
Ez
E x
0
cBz
cBy
E y
cBz
0
cB x
LNMM
E z
cBy
cB x O
0
(6.61)
F = A A =
IJJK
0
E x
E y
E z
Ex
0
cBz
cBy
Ey
cBz
0
cB x
E XAMPLE
6.1
LNMM
Ez
cBy
cB x O
0
(6.62)
where denotes the row number and the column number. Then, Einstein summation
82
I NTERACTIONS
OF
F IELDS
AND
PARTICLES
(6.63)
E 2x + 0 + c2 B2z + c2 B2y
E y2 + c2 B2z + 0 + c2 B2x
E z2 + c2 B2y + c2 B2x + 0
= 2E 2 + 2c2 B2 = 2(c2 B2 E 2 )
QED
E ND
OF EXAMPLE
6.1
=@
+@
1
= j A + 0 F F
(6.64)
4
inserted into the Euler-Lagrange equations, expression (6.49) on page 79, yields
two of Maxwells equations. To see this, we note from Equation (6.64) above
and the results in Example 6.1 that
EM
EM
inter
field
= j
(6.65)
Furthermore,
@
(
0
F F #
"
A
)
4
(
A
)
0
=
( A A )( A A )
4
( A )
EM
A A A A
4
( A )
(6.66)
A A + A A
=
A A A A #
2
( A ) "
6.2
83
But
A A # = A
A + A
A
( A ) "
( A )
( A )
= A
A + A
g g A
( A )
( A )
(6.67)
A + g g A
A
= A
( A )
( A )
= A
A + A
A
( A )
( A )
= 2 A
Similarly,
A A # = 2 A
( A ) "
(6.68)
so that
@
(
EM
A )
= 0 A A # = 0
"
F
x
(6.69)
This means that the Euler-Lagrange equations, expression (6.49) on page 79,
for the Lagrangian density @ EM and with A as the field quantity become
@
EM
@
(
EM
A )
= j 0
F
=0
x
(6.70)
or
F
j
=
x
0
(6.71)
E x Ey Ez
+
+
= E =
(6.72)
+
+
+
= 0+
0
1
2
3
x
x
x
x
x
y
z
0
which is the Maxwell source equation for the electric field, Equation (1.43a)
on page 14. For = 1 we get
By u x
1 E x
Bz
F 01 F 11 F 21 F 31
+
+
+
=
+0c
+c
=
(6.73)
0
1
2
3
x
x
x
x
c t
y
z
0
84
I NTERACTIONS
OF
F IELDS
AND
PARTICLES
0 0
= 0 j x
z
y
t
(6.74)
E
= 0 j(t, x)
t
(6.75)
which is the Maxwell source equation for the magnetic field, Equation (1.43d)
on page 14.
Other fields
In general, the dynamic equations for most any fields, and not only electromagnetic ones, can be derived from a Lagrangian density together with a variational principle (the Euler-Lagrange equations). Both linear and non-linear
fields are studied with this technique. As a simple example, consider a real,
scalar field which has the following Lagrange density:
1
1
m2 2 # =
2"
2
m2 2
x x
(6.76)
m2 ) = 0
(6.77)
em|x|
|x|
(6.78)
which describes the Yukawa meson field for a scalar meson with mass m. With
=
1
c2 t
(6.79)
1 2 2
& c + " # 2 + m2 2 '
2
(6.80)
6.2
85
B IBLIOGRAPHY
Another Lagrangian density which has attracted quite some interest is the
Proca Lagrangian
EM
=@
inter
+@
field
1
= j A + 0 F F + m2 A A
4
(6.81)
x
0
(6.82)
Bibliography
[1] A. O. BARUT, Dynamics and Classical Theory of Fields and Particles, Dover
Publications, Inc., New York, NY, 1980, ISBN 0-486-64038-8.
[2] V. L. G INZBURG, Applications of Electrodynamics in Theoretical Physics
and Astrophysics, Revised third ed., Gordon and Breach Science Publishers,
New York, London, Paris, Montreux, Tokyo and Melbourne, 1989, ISBN 288124-719-9.
[3] H. G OLDSTEIN, Classical Mechanics, second ed., Addison-Wesley Publishing
Company, Inc., Reading, MA . . . , 1981, ISBN 0-201-02918-9.
[4] W. T. G RANDY, Introduction to Electrodynamics and Radiation, Academic
Press, New York and London, 1970, ISBN 0-12-295250-2.
[5] L. D. L ANDAU , AND E. M. L IFSHITZ, The Classical Theory of Fields, fourth revised English ed., vol. 2 of Course of Theoretical Physics, Pergamon Press, Ltd.,
Oxford . . . , 1975, ISBN 0-08-025072-6.
[6] W. K. H. PANOFSKY, AND M. P HILLIPS, Classical Electricity and Magnetism,
second ed., Addison-Wesley Publishing Company, Inc., Reading, MA . . . , 1962,
ISBN 0-201-05702-6.
[7] J. J. S AKURAI, Advanced Quantum Mechanics, Addison-Wesley Publishing
Company, Inc., Reading, MA . . . , 1967, ISBN 0-201-06710-2.
86
I NTERACTIONS
OF
F IELDS
AND
PARTICLES
7
Interactions of
Fields and
Matter
The microscopic Maxwell equations (1.43) derived in Chapter 1 are valid on all
scales where a classical description is good. However, when macroscopic matter is present, it is sometimes convenient to use the corresponding macroscopic
Maxwell equations (in a statistical sense) in which auxiliary, derived fields are
introduced in order to incorporate effects of macroscopic matter when this is
immersed fully or partially in an electromagnetic field.
(x0 ) d3x0
(7.1)
where the is the charge density introduced in Equation (1.7) on page 4, the
electric dipole moment vector
p=
(7.2)
87
(7.3)
88
I NTERACTIONS
OF
F IELDS
AND
M ATTER
q
(x x0 )i
1
1
pi
+
2
40 |x x0 | |x x0 |
|x x0 |
1
Qi j
+
|x x0 |3
3 (x x0 )i (x x0 ) j 1
i j + . . .
2 |x x0 | |x x0 | 2
(7.4)
p (x) =
(7.5)
Using the expression Equation (M.71) on page 173 and applying the divergence theorem, we can rewrite this expression for the potential as follows:
1
p (x) =
40
1
=
40
P(x0 )
0 P(x0 ) 3 0
3 0
d
x
dx
|x x0 |
V
V |x x0 |
0
0
0
P(x ) n 2
P(x ) 3 0
d x
dx
0|
|x
x
S
V |x x0 |
0
(7.6)
where the first term, which describes the effects of the induced, non-cancelling
dipole moment on the surface of the volume, can be neglected, unless there is
7.1
E LECTRIC
89
1
40
0 P(x0 ) 3 0
dx
|x x0 |
(7.7)
Comparing this expression with expression Equation (3.3) on page 33 for the
electrostatic potential from a static charge distribution , we see that P(x)
has the characteristics of a charge density and that, to the lowest order, the
effective charge density becomes (x) P(x), in which the second term is a
polarisation term.
The version of Equation (1.7) on page 4 where true and polarisation
charges are separated thus becomes
E =
(x) P(x)
0
(7.8)
Rewriting this equation, and at the same time introducing the electric displacement vector (C/m2 )
D = 0 E + P
(7.9)
we obtain
(0 E + P) = D = true (x)
(7.10)
where true is the true charge density in the medium. This is one of Maxwells equations and is valid also for time varying fields. By introducing the
notation pol = P for the polarised charge density in the medium, and
total = true + pol for the total charge density, we can write down the following alternative version of Maxwells equation (7.23a) on page 91
E =
total (x)
0
(7.11)
Often, for low enough field strengths |E|, the linear and isotropic relationship between P and E
P = 0 E
(7.12)
90
I NTERACTIONS
OF
F IELDS
AND
M ATTER
on the preceding page but non-linear terms are important. In such a situation
the principle of superposition is no longer valid and non-linear effects such as
frequency conversion and mixing can be expected.
Inserting the approximation (7.12) into Equation (7.9) on the previous
page, we can write the latter
D = E
(7.13)
where, approximately,
= 0 (1 + )
(7.14)
1
2
(7.15)
7.3
E NERGY
91
AND MOMENTUM
volume, M. Via the definition of the vector potential one can show that the
magnetisation current and the magnetisation is simply related:
jm = M
(7.16)
In a stationary medium we therefore have a total current which is (approximately) the sum of the three currents enumerated above:
P
jtotal = jtrue +
+M
(7.17)
t
We then obtain the Maxwell equation
B = 0 jtrue +
P
+M
t
(7.18)
Moving the term M to the left hand side and introducing the magnetising
field (magnetic field intensity, Ampre-turn density) as
B
H=
M
(7.19)
0
and using the definition for D, Equation (7.9) on page 89, we can write this
Maxwell equation in the following form
D
E
P
= jtrue +
0
(7.20)
H = jtrue +
t
t
t
We may, in analogy with the electric case, introduce a magnetic susceptibility for the medium. Denoting it m , we can write
B
H=
(7.21)
where, approximately,
= 0 (1 + m )
(7.22)
E =
(7.23a)
(7.23b)
B
t
H = j(t, x) +
(7.23c)
D
t
(7.23d)
92
I NTERACTIONS
OF
F IELDS
AND
M ATTER
and are called Maxwells macroscopic equations. These equations are convenient to use in certain simple cases. Together with the boundary conditions and
the constitutive relations, they describe uniquely (but only approximately!) the
properties of the electric and magnetic fields in matter. We shall use them in the
following considerations on the energy and momentum of the electromagnetic
field and its interaction with matter.
(7.24)
Integration over the entire volume V and using Gausss theorem (the divergence theorem), we obtain
t
1
(H B + E D) d3x0 =
2
j E d3x0 +
(E H) dS0
(7.25)
(7.26)
j E d3x0 =
j2 3 0
dx
j EEMF d3x0
(7.27)
*
j EEMF d3x0 =
+-,
*
j2 3 0
dx +
+-, . t *
Joule heat
*
Field energy
(E H) dS
+-,
1
(E D + H B) d3x0
2
+/,
.
(7.28)
Radiated power
7.3
E NERGY
93
AND MOMENTUM
(7.29)
Ue =
(7.30)
(7.31)
where U e is the electric field energy, U m is the magnetic field energy, both
measured in J, and S is the Poynting vector (power flux), measured in W/m2 .
D
B
t
D
B
= E( D) + ( H) B
t
= E( D) B ( H)
B
(D B) + D
t
t
= E( D) B ( H)
(D B) D ( E) + H(
* +-, B. )
t
E + j B = ( D)E + H
(7.32)
=0
(D B)
t
One verifies easily that the ith vector components of the two terms in
square brackets in the right hand member of (7.32) can be expressed as
D
E
1
E
D
+
[E( D) D ( E)]i =
2
xi
xi
x j
1
E i D j E D i j
2
(7.33)
and
94
I NTERACTIONS
OF
1
B
H
[H( B) B ( H)]i =
H
+
B
2
xi
xi
x j
F IELDS
AND
M ATTER
1
Hi B j B H i j
2
(7.34)
respectively.
Using these two expressions in the ith component of Equation (7.32) on
the previous page and re-shuffling terms, we get
1
D
B
E
H
+ H
+ (D B)i
(E + j B)i E
D
B
xi
xi
xi
xi
t
2
1
1
=
E i D j E D i j + Hi B j H B i j
x j
2
2
(7.35)
Introducing the electric volume force Fev via its ith component
1
(Fev )i = (E + j B)i
2
D
E
B
H
E
D
+ H
B
xi
xi
xi
xi P
(7.36)
(7.37)
Fev + t (D B)
=
i
T i j
x j
(7.38)
(7.39)
B = m 0 H = H
(7.40)
T i j
m S
= Fev + 2
x j
c t
(7.41)
i
7.3
95
B IBLIOGRAPHY
where S is the Poynting vector defined in Equation (7.29) on page 93. Integration over the entire volume V yields
*
Fev d3x0 +
+-,
d
dt *
m 3 0
Sd x =
c2+-,
.
Field momentum
*
T n d2x0
+-,
(7.42)
Maxwell stress
which expresses the balance between the force on the matter, the rate of change
of the electromagnetic field momentum and the Maxwell stress. This equation
is called the momentum theorem in Maxwells theory.
In vacuum (7.42) becomes
(E + v B) d3x0 +
1 d
c2 dt
S d3x0 =
V
T n d2x0
(7.43)
or
d mech d field
p
+ p
=
dt
dt
T n d2x0
(7.44)
Bibliography
[1] W. K. H. PANOFSKY, AND M. P HILLIPS, Classical Electricity and Magnetism,
second ed., Addison-Wesley Publishing Company, Inc., Reading, MA . . . , 1962,
ISBN 0-201-05702-6.
96
I NTERACTIONS
OF
F IELDS
AND
M ATTER
8
Electromagnetic
Radiation
Brad (t, x) =
Erad (t, x) =
=
where
def
0
j(tret
, x0 )
it
Brad
d =
(x) e
it
Erad
d
(x) e
1
40 c2
0
4c
j
t
0 , x0 ) (x x0 )
j(tret
d3x0
|x x0 |2
0 , x0 ) (x x0 )] (x x0 )
[j(tret
d3x0
|x x0 |3
(8.1a)
(8.1b)
(8.2)
0
t=tret
Instead of studying the fields in the time domain, we can often make a
spectrum analysis into the frequency domain and study each Fourier compon-
97
98
E LECTROMAGNETIC R ADIATION
S
dS = nd
2x
k
x x0
x x0
x
x0
x0 x 0
x0
V
O
F IGURE 8.1: Relation between the surface normal and the k vector for
radiation generated at source points x0 near the point x0 in the source
volume V. At distances much larger than the extent of V, the unit vector
n,
normal to the surface S which has its centre at x0 , and the unit vector
k of the radiation k vector from x0 are nearly coincident.
Brad
(x) =
(8.3a)
(8.3b)
If the source is located inside a volume V near x0 and has such a limited
spatial extent that max |x0 x0 | |x x0 |, and the integration surface S , centred
on x0 , has a large enough radius |x x0 | max |x0 x0 |, we see from Figure 8.1
8.2
R ADIATED
99
ENERGY
k x x0 k (x x0 ) k (x x0 ) k (x0 x0 )
(8.4)
k |x x0 | k (x0 x0 )
Recalling from Formula (F.48) and formula (F.49) on page 154 that
dS = |x x0 |2 d = |x x0 |2 sin d d
and noting from Figure 8.1 on the preceding page that k and n are nearly
parallel, we see that we can approximate.
k dS
k n
=
dS d
|x x0 |2 |x x0 |2
(8.5)
i
e
dx
e
40 c
|x x0 |2
V
1 eik|xx0 | (x x0 )
0
i
[j (x0 ) k] eik(x x0 ) d3x0
40 c |x x0 | |x x0 | V
(8.6b)
Brad
(x) i
100
E LECTROMAGNETIC R ADIATION
hSi = hE Hi =
Using the far-field approximations (8.6a) and (8.6b) and the fact that 1/c =
(j k)e
ik(x0 x0 ) 3 0
dx
2
x x0
|x x0 |
(8.8)
1
dP
=
R0
2
d 32
(j k)e
ik(x0 x0 ) 3 0
dx
2
(8.9)
S(t) dt =
(E H) dt
d
(E H0 )e
i(+0 )t
(8.10)
dt
If we carry out the temporal integration first and use the fact that
ei(+ )t dt = 2( + 0 )
(8.11)
8.3
R ADIATION
101
Equation (8.10) on the preceding page can be written [cf. Parsevals identity]
S(t) dt = 2
= 2
= 2
(E H ) d +
(E H ) d
(E H ) d
2
0
2
=
0
= 2
(E H ) d
(E H ) d +
(E H ) d
(E H ) d
(8.12)
(E B + E B ) d
(E B + E B ) d
where the last step follows from the real-valuedness of E and B . We insert the Fourier transforms of the field components which dominate at large
distances, i.e., the radiation fields (8.3a) and (8.3b).
The result, after integration over the area S of a large sphere which encloses the source, is
1
U=
4
0
0
S
2
j k ik|xx0 | 3 0
e
d x k ndS
d
|x x0 |
(8.13)
Inserting the approximations (8.4) and (8.5) into Equation (8.13) and also
introducing
U=
U d
(8.14)
and recalling the definition (2.18) on page 26 for the vacuum resistance R 0 we
obtain
dU
1
d R0
d
4
(j k)e
ik(x0 x0 ) 3 0
dx
2
(8.15)
102
E LECTROMAGNETIC R ADIATION
j0 (x ) =
sin[k(L/2
I0 (x10 )(x20 )
0
x3 )]
sin(kL/2)
x 3
(8.16)
8.3
R ADIATION
103
the expression
(j0 k)e
=
dx
I02
2
2
sin[k(L/2 x30 )]
0
ikx03 cos ikx0 cos
e
dx3
I0
k sin e
sin(kL/2)
L/2
L/2
ik(x0 x0 ) 3 0
4I02
L/2
0
2
(8.17)
inserting this expression and d = 2 sin d into formula (8.9) on page 100
and integrating over , we find that the total radiated power from the antenna
is
P(L) =
R0 I02
1
4
sin d
(8.18)
lim P(L) =
kL0
12
L
R0 J02
(8.19)
P(L) P(L)
R (L) = 2 = 1 2 = R0
6
Ieff
2 I0
rad
L
197
L
(8.20)
1
4
cos2
" 2 cos
sin
(8.21)
104
E LECTROMAGNETIC R ADIATION
analytically as follows:
cos2
" 2 cos
sin
cos2 2 v#
" dv =
1 1 v2
Q
1 + cos(u)
cos2 v =
2
2
R
1
1 + cos(u)
1
=
du
2 1 (1 + u)(1 u)
(8.22)
1 1 1 + cos(u)
1 1 1 + cos(u)
du +
du
=
4 1 (1 + u)
4 1 (1 u)
1 1 1 + cos(u)
v
=
du = S 1 + u
2 1 (1 + u)
T
2
1 cos v
1
1
=
dv = [ + ln 2 Ci(2)]
2 0
v
2
1.22
1
d = dcos vc =
where in the last step the Euler-Mascheroni constant = 0.5772 . . . and the
cosine integral Ci(x) were introduced. Inserting this into the expression Equation (8.21) on the preceding page we obtain the value Rrad (/2) 73 .
(8.23)
= jtrue
t
(8.24a)
(8.24b)
8.4
M ULTIPOLE
105
RADIATION
and compare with Equation (8.23) on the facing page, we see that (t, x) satisfies this continuity equation. Furthermore, if we compare with the electric
polarisation [cf. Equation (7.9) on page 89], we see that the quantity is related to the true charges in the same way as P is related to polarised charge.
Therefore, is referred to as the polarisation vector.
We introduce a further potential e with the following property
e =
1 e
=A
c2 t
(8.25a)
(8.25b)
where and A are the electromagnetic scalar and vector potentials, respectively. As we see, e acts as a super-potential in the sense that it is a potential from which we can obtain other potentials. It is called the Hertz vector
or polarisation potential and, as can be seen from (8.24) and (8.25), it satisfies
the inhomogeneous wave equation
e =
1 2 e
2 e =
2
2
c t
0
(8.26)
This equation is of the same type as Equation (3.19) on page 38, and has
therefore the retarded solution
e (t, x) =
1
40
0 , x0 )
(tret
d3x0
|x x0 |
(8.27)
1
40
(x0 )eik|xx | 3 0
dx
|x x0 |
(8.28)
(8.29)
we see that we can calculate the magnetic and electric fields, respectively, as
follows
1 C
c2 t
(8.30a)
E = C
(8.30b)
B=
Clearly, the last equation is valid only outside the source volume, where
E = 0. Since we are mainly interested in the fields in the far zone, a long
distance from the source region, this is no essential limitation.
106
E LECTROMAGNETIC R ADIATION
Assume that the source region is a limited volume around some central
point x0 far away from the field (observation) point x. Under these assumptions, we can expand expression (8.27) on the previous page the Hertz vector,
0 , x0 ) in the vicinity of x , in a formal
due to the presence of non-vanishing (tret
0
series. For this purpose we recall from potential theory that
0
eik|xx |
eik|(xx0 )(x x0 )|
|x x0 | |(x x0 ) (x0 x0 )|
(8.31)
n=0
where
0
eik|xx |
is a Greens function
|x x0 |
is the angle between x0 x0 and x x0
h(1)
n (k |x x0 |) is the spherical Hankel function of the first kind of order n
According to the addition theorem for Legendre polynomials, we can write
n
Pn (cos ) =
0 im( )
(1)m Pmn (cos )Pm
n (cos )e
0
(8.32)
m=n
where Pm
n is an associated Legendre polynomial and, in spherical polar coordinates,
x0 x0 = ( x0 x0 , 0 , 0 )
(8.33a)
x x0 = (|x x0 | , , )
(8.33b)
Inserting Equation (8.31), together with Equation (8.32) above, into Equation (8.28) on the preceding page, we can in a formally exact way expand
the Fourier component of the Hertz vector as
ik n
m
im
(2n + 1)(1)mh(1)
n (k |x x0 |) Pn (cos ) e
40 n=0
m=n
e =
(x ) jn (k x
(8.34)
0 im0 3 0
x0 ) Pm
dx
n (cos ) e
We notice that there is no dependence on x x0 inside the integral; the integrand is only dependent on the relative source vector x0 x0 .
8.4
M ULTIPOLE
107
RADIATION
We are interested in the case where the field point is many wavelengths
away from the well-localised sources, i.e., when the following inequalities
k x0 x0 1 k |x x0 |
(8.35)
eik|xx0 |
k |x x0 |
(8.36)
and replace jn with the first term in its power series expansion:
jn (k x0 x0 )
2n n!
k x0 x0 #
(2n + 1)! "
(8.37)
Inserting these expansions into Equation (8.34) on the preceding page, we obtain the multipole expansion of the Fourier component of the Hertz vector
e
e(n)
(8.38a)
n=0
where
e(n) = (i)n
1 eik|xx0 | 2n n!
40 |x x0 | (2n)!
108
E LECTROMAGNETIC R ADIATION
eik|xx0 |
40 |x x0 |
(x0 ) d3x0 =
1 eik|xx0 |
p
40 |x x0 |
(8.39)
where p = V (x0 ) d3x0 is the Fourier component of the electric dipole moment; cf. Equation (7.2) on page 87 which describes the static dipole moment.
If a spherical coordinate system is chosen with its polar axis along p , the
components of e (0) are
1 eik|xx0 |
p cos
40 |x x0 |
1 eik|xx0 |
= (0)
p sin
sin =
40 |x x0 |
= 0
r = (0)
cos =
(8.40a)
(8.40b)
(8.40c)
Evaluating formula (8.29) on page 105 for the help vector C, with the
spherically polar components (8.40) of e (0)
inserted, we obtain
(0)
C = C,
1
=
40
1
eik|xx0 |
ik
|x x0 |
|x x0 | p sin
(8.41)
Applying this to Equation (8.30) on page 105, we obtain directly the Fourier
components of the fields
0
1
eik|xx0 |
B = i
ik
p sin
4 |x x0 |
|x x0 |
1
ik
1
x x0
E =
2
cos
2
4
|x x0 |
|x x0 |
|x x0 |
0
ik|xx
0|
1
ik
2
e
k
sin
|x x0 | p
|x x0 |2 |x x0 |
(8.42a)
(8.42b)
Keeping only those parts of the fields which dominate at large distances
(the radiation fields) and recalling that the wave vector k = k(x x0 )/ |x x0 |
where k = /c, we can now write down the Fourier components of the radiation
parts of the magnetic and electric fields from the dipole:
0 eik|xx0 |
(p k)
4 |x x0 |
1 eik|xx0 |
Erad
[(p k) k]
=
40 |x x0 |
Brad
=
(8.43a)
(8.43b)
8.4
M ULTIPOLE
109
RADIATION
These fields constitute the electric dipole radiation, also known as E1 radiation.
eik|xx0 |
k x0 x0 (x0 ) cos d3x0
40 |x x0 | V
1 eik|xx0 |
[(x x0 ) (x0 x0 )] (x0 ) d3x0
= ik
40 |x x0 |2 V
e(1) = i
(8.44)
(8.45)
and introducing
i = xi x0,i
0i
(8.46a)
xi0 x0,i
(8.46b)
(8.47)
i.e., as the sum of two parts, the first being symmetric and the second antisymmetric in the indices i, j. We note that the antisymmetric part can be written
as
1
1
i , j 0i ,i 0j # = [, j (i 0i ) 0j (i , j )]
2 "
2
1
= [ ( 0 ) 0 ( )] j
2
1
(x x0 ) [ (x0 x0 )] !
=
2
(8.48)
j
Equations (8.24) on page 104, and the fact that we are considering a single
Fourier component,
(t, x) = eit
(8.49)
110
E LECTROMAGNETIC R ADIATION
allow us to express in j as
= i
(8.50)
Hence, we can write the antisymmetric part of the integral in formula (8.44)
on the preceding page as
1
(x x0 ) (x0 ) (x0 x0 ) dV 0
2
V
1
= i (x x0 ) j (x0 ) (x0 x0 ) d3x0
2
V
1
= i (x x0 ) m
(8.51)
1
2
(8.52)
The final result is that the antisymmetric, magnetic dipole, part of e(1)
can be written
e,antisym
(1)
k eik|xx0 |
(x x0 ) m
40 |x x0 |2
(8.53)
In analogy with the electric dipole case, we insert this expression into Equation (8.29) on page 105 to evaluate C, with which Equations (8.30) on page 105
then gives the B and E fields. Discarding, as before, all terms belonging to the
near fields and transition fields and keeping only the terms that dominate at
large distances, we obtain
0 eik|xx0 |
(m k) k
4 |x x0 |
k eik|xx0 |
m k
Erad
(x) =
40 c |x x0 |
Brad
(x) =
(8.54a)
(8.54b)
which are the fields of the magnetic dipole radiation (M1 radiation).
8.5
R ADIATION
111
of the electric quadrupole tensor, which is defined in accordance with Equation (7.3) on page 87:
Q(t) =
(8.55)
Again we use this expression in Equation (8.29) on page 105 to calculate the
fields via Equations (8.30) on page 105. Tedious, but fairly straightforward
algebra (which we will not present here), yields the resulting fields. The radiation components of the fields in the far field zone (wave zone) are given
by
i0 eik|xx0 |
k Q # k
8 |x x0 | "
i eik|xx0 |
k Q # k k
Erad
(x) =
80 |x x0 | "
Brad
(x) =
(8.56a)
(8.56b)
(t, x) =
(8.57a)
(8.57b)
and consider a source region with such a limited spatial extent that the charges
and currents are well localised. Specifically, we consider a charge q0 , for instance an electron, which, classically, can be thought of as a localised, unstructured and rigid charge distribution with a small, finite radius. The part of this
charge distribution dq0 which we are considering is located in dV 0 = d3x0 in
the sphere in Figure 8.5 on the following page. Since we assume that the electron (or any other other similar electric charge) is moving with a velocity v
whose direction is arbitrary and whose magnitude can be almost comparable
to the speed of light, we cannot say that the charge and current to be used in
0 , x0 ) d3x0 and
0 , x0 ) d3x0 , respectively, because in the fi(8.57) is V (tret
V v(tret
nite time interval during which the observed signal is generated, part of the
charge distribution will leak out of the volume element d3x0 .
112
E LECTROMAGNETIC R ADIATION
x(t)
dr
q0
dV 0
F IGURE 8.2: Signals which are observed at the field point x at time t
were generated at source points x0 (t0 ) on a sphere, centred on x and expanding, as time increases, with the velocity c outward from the centre.
The source charge element moves with an arbitrary velocity v and gives
rise to a source leakage out of the source volume dV 0 = d3x0 .
(x x0 ) v
dS dt
|x x0 |
(8.58)
where the last term represents the amount of source leakage due to the fact
that the charge distribution moves with velocity v. Since dt = dr/c and dS dr =
d3x0 we can rewrite this expression for the net charge as
(x x0 ) v 3 0
dx
c |x x0 |
(x x0 ) v
0
0
= (tret , x ) 1
d3x0
c |x x0 |
0
0
dq0 = (tret
, x0 ) d3x0 (tret
, x0 )
(8.59)
8.5
R ADIATION
113
or
0
(tret
, x0 ) d3x0 =
dq0
(8.60)
)v
1 (xx
c|xx0 |
(8.61)
This is the expression to be used in the formulae (8.57) on page 111 for the
retarded potentials. The result is (recall that j = v)
(t, x) =
1
40
A(t, x) =
0
4
dq0
|x x0 | (xxc0 )v
v dq0
0
|x x0 | (xxc )v
(8.62a)
(8.62b)
For a sufficiently small and well localised charge distribution we can, assuming
that the integrands do not change sign in the integration volume, use the mean
value theorem and the fact that V dq0 = q0 to evaluate these expressions to
become
1
q0
q0 1
=
0
40 |x x0 | (xxc )v 40 s
q0
v
q0 v v
A(t, x) =
= (t, x)
=
0
40 c2 |x x0 | (xxc )v 40 c2 s c2
(t, x) =
(8.63a)
(8.63b)
where
s = x x0
(x x0 ) v
c
0
x x0 v
= xx 1
|x x0 | c
x x0 v
0
= (x x )
|x x0 | c
(8.64a)
(8.64b)
(8.64c)
is the retarded relative distance. The potentials (8.63) are precisely the LinardWiechert potentials which we derived in Section 5.3.2 on page 63 by using a
covariant formalism.
It is important to realise that in the complicated derivation presented here,
the observer is in a coordinate system which has an absolute meaning and the
114
E LECTROMAGNETIC R ADIATION
?
|x x0 |
v
c
q0
x0 (t0 )
v(t0 )
x0 (t)
x x0
x x0
x(t)
F IGURE 8.3: Signals which are observed at the field point x at time t
were generated at the source point x0 (t0 ). After time t0 the particle, which
moves with nonuniform velocity, has followed a yet unknown trajectory.
Extrapolating tangentially the trajectory from x0 (t0 ), based on the velocity v(t0 ), defines the virtual simultaneous coordinate x0 (t).
velocity v is that of the particle, whereas in the covariant derivation two frames
of equal standing were moving relative to each other with v. Expressed in the
four-potential, Equation (5.45) on page 62, the Linard-Wiechert potentials
become
q0
A (x ) =
40
1 v
,
= (, cA)
s cs
(8.65)
(8.66a)
A(t, x)
E(t, x) = (t, x)
t
(8.66b)
(8.67)
8.5
R ADIATION
115
(in the interest of simplifying our notation, we drop the subscript ret on t 0
from now on). This means that we know the trajectory of the charge q0 , i.e., x0 ,
for all times up to the time t0 at which a signal was emitted in order to precisely
arrive at the field point x at time t. Because of the finite speed of propagation
of the fields, the trajectory at times later than t 0 is not (yet) known.
The retarded velocity and acceleration at time t 0 are given by
v(t0 ) =
dx0
dt0
a(t0 ) = v (t0 ) =
(8.68a)
dv d2 x0
=
dt0 dt0 2
(8.68b)
(8.69a)
d2
(x x0 (t0 )) = v(t0 )
dt0 2
(8.69b)
The retarded time t0 can, at least in principle, be calculated from the implicit
relation
t0 = t0 (t, x) = t
|x x0 (t0 )|
c
(8.70)
and we shall see later how this relation can be taken into account in the calculations.
According to formulae (8.66) on the facing page the electric and magnetic
fields are determined via differentiation of the retarded potentials at the observation time t and at the observation point x. In these formulae the unprimed
, i.e., the spatial derivative differentiation operator = x i /xi means that
we differentiate with respect to the coordinates x = (x1 , x2 , x3 ) while keeping t
fixed, and the unprimed time derivative operator /t means that we differentiate with respect to t while keeping x fixed. But the Linard-Wiechert potentials
and A, Equations (8.63) on page 113, are expressed in the charge velocity
v(t0 ) given by Equation (8.68a) and the retarded relative distance s(t 0 , x) given
by Equation (8.64) on page 113. This means that the expressions for the potentials and A contain terms which are expressed explicitly in t 0 , which in turn
is expressed implicitly in t via Equation (8.70). Despite this complication it is
116
E LECTROMAGNETIC R ADIATION
possible, as we shall see below, to determine the electric and magnetic fields
and associated quantities at the time of observation t. To this end, we need to
investigate carefully the action of differentiation on the potentials.
The differential operator method We introduce the convention that a differential operator embraced by parentheses with an index x or t means that
the operator in question is applied at constant x and t, respectively. With this
convention, we find that
t0
x
x x0
x x (t ) =
0
|x x | t0
0 0
"
x x0 (t0 )# =
(x x0 ) v
|x x0 |
(8.71)
Furthermore, by applying the operator (/t)x to Equation (8.70) on the previous page we find that
t0
t
t
|x x0 |
= 1
0
t x c
0 t
0
(x x ) v t
= 1+
c |x x0 |
t x
= 1
(8.72)
x
t0
t
|x x0 |
|x x0 |
=
s
|x x0 | (x x0 ) v/c
=
x
(8.73)
t
=
x
t0
t
t0
|x x0 |
=
s
x
t0
(8.74)
x
()t t0 = ()t
(8.75)
8.5
R ADIATION
117
x x0
cs
(8.76)
which gives the following operator relation when ()t is acting on an arbitrary
function of t0 and x:
()t = ()t t
t0
x x0
+ ()t0 =
cs
x
t0
(8.77)
+ ()t0
x
With the help of the rules (8.77) and (8.74) we are now able to replace t by t 0
in the operations which we need to perform. We find, for instance, that
1 q0
40 s
x x0 v x x0 s
q0
=
40 s2 |x x0 | c
cs
t0 x
A
A
0 q0 v
=
t
t x t 4 s
0
0 s
q0
=
x x sv x x v t0
40 c2 s3
()t =
(8.78a)
(8.78b)
x
Utilising these relations in the calculation of the E field from the LinardWiechert potentials, Equations (8.63) on page 113, we obtain
E(t, x) =
A
t
q0
(x x0 ) |x x0 | v/c
40 s2
|x x0 |
(x x0 ) |x x0 | v/c
cs
(8.79)
s
t0
|x x0 | v
c2
Starting from expression (8.64a) on page 113 for the retarded relative distance
s(t0 , x), we see that we can evaluate (s/t 0 )x in the following way
s
t0
=
x
t0
0
x x0 (x x ) v
1
x x0 (t0 )
0
t
c
"
x x0 (t0 )
t0
v (x x0 )
v(t0 )
t0 1
(8.80)
(x x0 ) v v2 (x x0 ) v
+
c
c
|x x0 |
118
E LECTROMAGNETIC R ADIATION
where Equation (8.71) on page 116 and Equations (8.68) on page 115, respectively, were used. Hence, the electric field generated by an arbitrarily moving
charged particle at x0 (t0 ) is given by the expression
q0
E(t, x) =
40 s3
|x x0 | v
(x x )
+-c,
0
v2
c2 .
x x0
+ 2
* c
|x x0 | v
(x x )
c
v .
+/,
0
(8.81)
Radiation field
The first part of the field, the velocity field, tends to the ordinary Coulomb field
when v 0 and does not contribute to the radiation. The second part of the
field, the acceleration field, is radiated into the far zone and is therefore also
called the radiation field.
From Figure 8.5.2 on page 114 we see that the position the charged particle
would have had if at t0 all external forces would have been switched off so that
the trajectory from then on would have been a straight line in the direction of
the tangent at x0 (t0 ) is x0 (t), the virtual simultaneous coordinate. During the
arbitrary motion, we interpret x x0 as the coordinate of the field point x relative to the virtual simultaneous coordinate x0 (t). Since the time it takes from a
signal to propagate (in the assumed vacuum) from x0 (t0 ) to x is |x x0 | /c, this
relative vector is given by
|x x0 | v
c
This allows us to rewrite Equation (8.81) in the following way
x x0 = (x x0 )
(8.82)
q0
v2
(x x0 ) v
E(t, x) =
(x
x
)
1
+ (x x0 )
0
3
2
40 s
c
c2
(8.83)
x x0
B(t, x) = A ()t A = () A
cs
t0
q0
x x0
A
x x0
=
v
0
2
2
0
40 c s |x x |
c |x x |
t x
t0
(8.84)
where we made use of Equation (8.63) on page 113 and formula (8.74) on
page 116. But, according to (8.78a),
x x0
x x0
q0
()
=
v
t
c |x x0 |
40 c2 s2 |x x0 |
(8.85)
8.5
R ADIATION
119
so that
x x0
A
B(t, x) =
()t
0
c |x x |
t
0
xx
=
E(t, x)
c |x x0 |
(8.86)
The radiation part of the electric field is obtained from the acceleration
field in formula (8.81) on the preceding page as
Erad (t, x) =
=
lim E(t, x)
|xx0 |
q0
(x x )
3
(x x0 )
40 c2 s
q0
=
(x x0 ) [(x x0 ) v ]
40 c2 s3
|x x0 | v
c
v
(8.87)
where in the last step we again used formula (8.82) on the facing page. Using
this formula and formula (8.86) above, the radiation part of the magnetic field
can be written
Brad (t, x) =
x x0
Erad (t, x)
c |x x0 |
(8.88)
s (t , x) = x x
(x x0 ) v
(x x0 ) v
+
2 xx
c
c
0 2
(8.89)
(x x0 ) v
c
(x x0 ) v
c
02 2
|x x0 |2 v2
2 0 |x x | v
cos
+
sin2 0
c2
c2
|x x0 |2 v2
|x x0 |2 v2
2 0
2 0
=
(cos
+
sin
)
=
c2
c2
(8.90)
120
E LECTROMAGNETIC R ADIATION
we find that
(x x0 ) v
c
(x x0 ) v
|x x0 |2 v2
=
c2
c
(8.91)
Furthermore, from Equation (8.82) on page 118, we obtain the following identity:
(x x0 ) v = (x x0 ) v
(8.92)
(x x0 ) v
c
(x x0 ) v
|x x0 |2 v2
=
c2
c
(8.93)
Inserting the above into expression (8.89) on the previous page for s2 , this
expression becomes
2
s = xx
0 2
(x x0 ) v |x x0 |2 v2
(x x0 ) v
2 xx
+
2
c
c
c
0
|x x0 | v
= (x x )
c
0
(x x0 ) v
c
(x x0 (t)) v
= (x x0 (t))
c
(8.94)
|x x0 |
(x x0 ) v
c
8.5
R ADIATION
121
Taking the square root of both sides of Equation (8.94) on the preceding
page, we obtain the following alternative final expressions for the retarded relative distance s in terms of the charges virtual simultaneous coordinate x 0 (t):
s(t, x) =
|x x0 |
= |x x0 | 1
(x x0 ) v
c
|x x0 |
v2 2
sin 0
c2
(8.95a)
(8.95b)
(x x0 ) v
v2
1 2
c
c
(8.95c)
v2
(x x0 ) v
1 2
c
c
v2
vv
+ 2 (x x0 )
2
c
c
v
v
= 2 & (x x0 ) + (x x0 ))'
c
c
= 2 (x x0 ) 1
;
(8.96)
T HE FIELDS FROM A
In the special case of uniform motion, the localised charge moves in a field-free, isolated space and we know that it will not be affected by any external forces. It will
therefore move uniformly in a straight line with the constant velocity v. This gives us
the possibility to extrapolate its position at the observation time, x0 (t), from its position at the retarded time, x0 (t0 ). Since the particle is not accelerated, v 0, the virtual
simultaneous coordinate x0 will be identical to the actual simultaneous coordinate of
the particle at time t, i.e., x0 (t) = x0 (t). As depicted in Figure 8.5.2 on page 114, the
angle between x x0 and v is 0 while then angle between x x0 and v is 0 .
E XAMPLE
8.1
We note that in the case of uniform velocity v, time and space derivatives are closely
related in the following way when they operate on functions of x(t) :
v
t
(8.97)
Hence, the E and B fields can be obtained from formulae (8.66) on page 114, with the
122
E LECTROMAGNETIC R ADIATION
s2
40 s
80 s3
(8.99)
q0
v
v
=
\ (x x0 ) + c Z c (x x0 )[ ]
40 s3
When this expression for is inserted into Equation (8.98a), the following result
vv
vv
q0
E(t, x) = 2 1[ =
1 [ s2
80 s3 Z c2
Zc
v
v
q0
(x x0 ) +
=
(x x0 )[
40 s3 ^
c Z c
v v
vv v
v
(x x0 ) [ 2 \
(x x0 )[
c Z c
c
c Z c
q0
v v
v2
[
(x
x
)
+
(x
x
)
(x
x
)
0
0
0
40 s3 a
c Z c
c2
]`_
(8.100)
v v
(x x0 ) [cb
c Z c
q0
v2
1
(x
x
)
0
40 s3
c2
follows. Of course, the same result also follows from Equation (8.83) on page 118
with v 0 inserted.
=
From Equation (8.100) we conclude that E is directed along the vector from the simultaneous coordinate x0 (t) to the field (observation) coordinate x(t). In a similar way,
the magnetic field can be calculated and one finds that
B(t, x) =
0 q 0
4s3
v2
v (x x0 ) = c12 v E
c2
(8.101)
8.5
R ADIATION
123
From these explicit formulae for the E and B fields we can discern the following cases:
1. v 0 E goes over into the Coulomb field ECoulomb
2. v 0 B goes over into the Biot-Savart field
3. v c E becomes dependent on 0
4. v c, sin 0 0 E (1 v2 /c2 )ECoulomb
5. v c, sin 0 1 E (1 v2 /c2 )1/2 ECoulomb
E ND OF
EXAMPLE
8.1
Let us consider in more detail the treatment of the radiation from a uniformly moving
rigid charge distribution.
E XAMPLE
8.2
If we return to the original definition of the potentials and the inhomogeneous wave
equation, formula (3.19) on page 38, for a generic potential component (t, x) and a
generic source component f (t, x),
(t, x) =
1 2
2 (t, x) = f (t, x)
c2 t2
(8.102)
we find that under the assumption that v = v x 1 , this equation can be written
v2 2 2 2
+
+
= f (x)
c2 x21 x22 x23
(8.103)
2 = x 2
3 = x 3
x1
(8.104a)
1 v2 /c2
(8.104b)
(8.104c)
def
124
E LECTROMAGNETIC R ADIATION
1 v2 /c2 1 , 2 , 3 ) f ()
(8.105)
in this space. This equation has the well-known Coulomb potential solution
() =
1
4 f
f ( 0 ) 3 0
0 d
(8.106)
(8.107)
(x) =
s=
(x1 x01 )2 +
v2
1 2 [(x2 x02 )2 + (x3 x03 )2 ]b
c
1
2
(8.108)
Applying this to the explicit scalar and vector potential components, realising that for
a rigid charge distribution moving with velocity v the current is given by j = v, we
obtain
(x0 ) 3 0
1
dx
(8.109a)
(t, x) =
40 f V s
1
v(x0 ) 3 0 v
A(t, x) =
d x = 2 (t, x)
(8.109b)
2
40 c f V s
c
For a localised charge where g d3x0 = q0 , these expressions reduce to
q0
40 s
q0 v
A(t, x) =
40 c2 s
(t, x) =
(8.110a)
(8.110b)
(8.111)
[ =
[
c Z c
c c2
Zc
(8.113)
8.5
R ADIATION
125
(8.114)
where
= 1
v2
c2
(8.115)
The scalar function is called the convection potential or the Heaviside potential.
When the rigid charge distribution is well localised so that we can use the potentials
(8.110) the convection potential becomes
= 1
v2
q0
2
c
40 s
(8.116)
x1 x01
1 v2 /c2
(8.117)
EXAMPLE
8.2
s = x x0
(x x0 ) v
x x0 ,
c
vc
(8.118)
|x x0 | v
x x0 ,
c
vc
(8.119)
126
E LECTROMAGNETIC R ADIATION
so that the radiation field Equation (8.87) on page 119 can be approximated by
Erad (t, x) =
q0
(x x0 ) [(x x0 ) v ],
40 c2 |x x0 |3
vc
(8.120)
from which we obtain, with the use of formula (8.86) on page 119, the magnetic field
Brad (t, x) =
q0
[v (x x0 )],
4c3 |x x0 |2
vc
(8.121)
(8.122)
x x = x x0
(8.123a)
(8.123b)
The power flux in the far zone is described by the Poynting vector as a
function of Erad and Brad . We use the close correspondence with the dipole
case to find that it becomes
S=
0
0 q0 2 (v)2
2 xx
sin
|x x0 |
162 c |x x0 |2
(8.124)
where is the angle between v and x x0 . The total radiated power (integrated
over a closed spherical surface) becomes
P=
0 q0 2 (v)2
q0 2 v 2
=
6c
60 c3
(8.125)
which is the Larmor formula for radiated power from an accelerated charge.
Note that here we are treating a charge with v c but otherwise totally unspecified motion while we compare with formulae derived for a stationary oscillating dipole. The electric and magnetic fields, Equation (8.120) and Equation (8.121) above, respectively, and the expressions for the Poynting flux and
power derived from them, are here instantaneous values, dependent on the instantaneous position of the charge at x0 (t0 ). The angular distribution is that
which is frozen to the point from which the energy is radiated.
8.5
R ADIATION
127
8.5.3 Bremsstrahlung
An important special case of radiation is when the velocity v and the acceleration v are collinear (parallel or anti-parallel) so that v v = 0. This condition
(for an arbitrary magnitude of v) inserted into expression (8.87) on page 119
for the radiation field, yields
Erad (t, x) =
q0
(x x0 ) [(x x0 ) v ],
40 c2 s3
v k v
(8.126)
from which we obtain, with the use of formula (8.86) on page 119, the magnetic field
Brad (t, x) =
q0 |x x0 |
[v (x x0 )],
40 c3 s3
v k v
(8.127)
The difference between this case and the previous case of v c is that the
approximate expression (8.118) on page 125 for s is no longer valid; we must
instead use the correct expression (8.64) on page 113. The angular distribution
of the power flux (Poynting vector) therefore becomes
S=
sin2
0 q0 2 v 2
162 c |x x0 |2 1 v cos #
c
"
x x0
|x x0 |
(8.128)
It is interesting to note that the magnitudes of the electric and magnetic fields
are the same whether v and v are parallel or anti-parallel.
We must be careful when we compute the energy (S integrated over time).
The Poynting vector is related to the time t when it is measured and to a fixed
surface in space. The radiated power into a solid angle element d, measured
relative to the particles retarded position, is given by the formula
0 q0 2 v 2
sin2
dU rad ()
d = S (x x0 ) x x0 d =
dt
162 c 1 v cos #
c
"
d (8.129)
On the other hand, the radiation loss due to radiation from the charge at retarded time t0 :
dU rad
dU rad
d
=
dt0
dt
t
t0
(8.130)
(8.131)
128
E LECTROMAGNETIC R ADIATION
v = 0.5c
v = 0.25c
v=0
v
F IGURE 8.4: Polar diagram of the energy loss angular distribution factor
sin2 /(1 v cos /c)5 during bremsstrahlung for particle speeds v = 0, v =
0.25c, and v = 0.5c.
"
(8.132)
The angular factors of this expression, for three different particle speeds, are
plotted in Figure 8.5.3.
Comparing expression (8.129) on the previous page with expression (8.132),
we see that they differ by a factor 1 v cos /c which comes from the extra
factor s/ |x x0 | introduced in (8.131). Let us explain this in geometrical terms.
During the interval (t0 , t0 + dt0 ) and within the solid angle element d the
particle radiates an energy [dU rad ()/dt0 ] dt0 d. As shown in 8.5.3 this energy
is at time t located between two spheres, one outer with its origin in x01 (t0 ) and
one inner with its origin in x01 (t0 + dt0 ) = x01 (t0 ) + v dt and radius c[t (t0 + dt0 )] =
c(t t0 dt0 ).
From Figure 8.5.3 we see that the volume element subtending the solid
angle element
d =
dS
2
x x02
is
(8.133)
2
d3x = dS dr = x x02 d dr
(8.134)
8.5
R ADIATION
129
dS
dr
x
d
0
q
0
x02 vdt x01
x x02 + c dt0
Here, dr denotes the differential distance between the two spheres and can be
evaluated in the following way
0
x x02 v dt0 x x02
* x +/x, 2 .
dr = x x02 + c dt0
v cos
=
x x02
x x02
dt0 =
(8.135)
cs
dt0
x x02
where formula (8.64) on page 113 was used in the last step. Hence, the volume
element under consideration is
d3x = dS dr =
s
dS c dt0
x x02
(8.136)
We see that the energy which is radiated per unit solid angle during the time
interval (t0 , t0 + dt0 ) is located in a volume element whose size is dependent. This explains the difference between expression (8.129) on page 127 and
expression (8.132) on the preceding page.
Let the radiated energy, integrated over , be denoted U rad . After tedious,
but relatively straightforward integration of formula (8.132) on the facing page,
130
E LECTROMAGNETIC R ADIATION
one obtains
dU rad 0 q0 2 v 2
=
dt0
6c
1 cv
2
2
(8.137)
If we know v(t0 ), we can integrate this expression over t 0 and obtain the total
energy radiated during the acceleration or deceleration of the particle. This
way we obtain a classical picture of bremsstrahlung (braking radiation). Often, an atomistic treatment is required for an acceptable result.
E XAMPLE
8.3
(8.138)
f v dt
(8.139)
(8.140)
(8.141)
From the general expression (8.86) on page 119 we conclude that E B and that
it suffices to consider E Erad . According to the bremsstrahlung expression for
Erad , Equation (8.126) on page 127,
E=
q0 sin
v (t0 t0 )
40 c2 |x x0 |
(8.142)
(8.143)
82
8.5
R ADIATION
U rad =
dU rad 0
dt =
f
dt0
131
1
=
0
f
EB dt0 d2x =
f S f
= 0 c
f S f E
E
1
0 c f
B
dS dt0
0
f E
dt0 d2x
(8.145)
dt0 d2x
According to Parsevals identity [cf. Equation (8.12) on page 101] the following equality holds:
E 2 dt0 = 4
f0
|E |2 d
(8.146)
|E |
fS
d2x d
(8.147)
For our infinite spectrum, Equation (8.144) on the preceding page, we obtain
q0 2 (v)2
U rad d =
163 0 c3
=
02
q (v)2
163 0 c3
q0 2
30 c
fS
f0
v
c
sin2 2
d x d
|x x0 |2
d
2
f0
sin2 sin d d
(8.148)
d
2
We see that the energy spectrum U rad is independent of frequency . This means that
if we integrate it over all frequencies [0, ], a divergent integral would result.
=
In reality, all spectra have finite widths, with an upper cutoff limit set by the quantum
condition
1
h = m(v)2
(8.149)
2
which expresses that the highest possible frequency in the spectrum is that for which
all kinetic energy difference has gone into one single field quantum (photon) with
energy h . If we adopt the picture that the total energy is quantised in terms of N
photons radiated during the process, we find that
U rad d
= dN
(8.150)
h
or, for an electron where q0 = |e|, where e is the elementary charge,
dN =
e2
2
40 h c 3
v
c
d
1 2
137 3
v
c
(8.151)
where we used the value of the fine structure constant e2 /(40 h c) 1/137.
Even if the number of photons becomes infinite when 0, these photons have
negligible energies so that the total radiated energy is still finite.
E ND OF
EXAMPLE
8.3
132
E LECTROMAGNETIC R ADIATION
(8.152a)
0 0
x (t ) = a[ x 1 cos (t ) + x 2 sin (t )]
v(t0 ) = x 0 (t0 ) = a0 [ x 1 sin (t0 ) + x 2 cos (t0 )]
v = |v| = a0
(8.152c)
(8.152d)
(8.152b)
a20
(8.152e)
(8.152f)
x x0 = x x0 ( x 2 sin + x 3 cos )
(8.153)
where is the angle between x x0 and the normal to the plane of the particle
orbit (see Figure 8.5.4). From the above expressions we obtain
(x x0 ) v = x x0 v sin cos
0
(8.154a)
(8.154b)
where in the last step we simply used the definition of a scalar product and the
fact that the angle between v and x x0 is .
The power flux is given by the Poynting vector, which, with the help of
formula (8.86) on page 119, can be written
S=
1
1
x x0
(E B) =
|E|2
0
c0
|x x0 |
(8.155)
(8.156)
8.5
R ADIATION
133
x2
(t, x)
x x0
x
v
q0 0 0
(t , x )
v
(t0 )
x1
x3
F IGURE 8.6: Coordinate system for the radiation from a charged particle
at x0 (t0 ) in circular motion with velocity v(t 0 ) along the tangent and constant acceleration v (t0 ) toward the origin. The x1 x2 axes are chosen so
that the relative field point vector x x0 makes an angle with the x3 axis
which is normal to the plane of the orbital motion. The radius of the orbit
is a.
where the retarded distance s is given by expression (8.64) on page 113. With
the radiation part of the electric field, expression (8.87) on page 119, inserted, and using (8.154a) and (8.154b) on the facing page, one finds, after some
algebra, that
#
2
2
v
v
dU rad (, ) 0 q0 2 v 2 " 1 c sin cos 1 c2 sin sin
(8.157)
=
5
dt0
162 c
1 v sin cos #
2
"
The angles and vary in time during the rotation, so that refers to a moving
coordinate system. But we can parametrise the solid angle d in the angle
and the (fixed) angle so that d = sin d d. Integration of Equation (8.157)
above over this d gives, after some cumbersome algebra, the angular integrated expression
dU rad 0 q0 2 v 2
=
dt0
6c
1 vc
2
2
(8.158)
134
E LECTROMAGNETIC R ADIATION
In Equation (8.157) on the preceding page, two limits are particularly interesting:
1. v/c 1 which corresponds to cyclotron radiation.
Cyclotron radiation
For a non-relativistic speed v c, Equation (8.157) on the previous page reduces to
dU rad (, ) 0 q0 2 v 2
=
(1 sin2 sin2 )
dt0
162 c
(8.159)
(8.160)
where is defined in Figure 8.5.4 on the previous page. This means that we
can write
0 q0 2 v 2
dU rad () 0 q0 2 v 2
2
=
(1
cos
)
=
sin2
dt0
162 c
162 c
(8.161)
Consequently, a fixed observer near the orbit plane will observe cyclotron
radiation twice per revolution in the form of two equally broad pulses of radiation with alternating polarisation.
Synchrotron radiation
When the particle is relativistic, v j c, the denominator in Equation (8.157)
on the preceding page becomes very small if sin cos 1, which defines the
forward direction of the particle motion ( /2, 0). Equation (8.157) on
the previous page then becomes
dU rad (/2, 0) 0 q0 2 v 2
1
=
0
2
dt
16 c 1 v #
c
"
(8.162)
which means that an observer near the orbit plane sees a very strong pulse
followed, half an orbit period later, by a much weaker pulse.
The two cases represented by Equation (8.161) and Equation (8.162) above
are very important results since they can be used to determine the characteristics of the particle motion both in particle accelerators and in astrophysical
objects where a direct measurement of particle velocities are impossible.
8.5
R ADIATION
135
x2
(t, x)
x x0
v
q0 0 0
(t , x )
v
(t0 )
0
x1
x3
F IGURE 8.7: When the observation point is in the plane of the particle
orbit, i.e., = /2 the lobe width is given by .
#
2
v
v
dU rad (/2, ) 0 q0 2 v 2 " 1 c cos 1 c2 sin
=
5
dt0
162 c
1 v cos #
2
"
(8.163)
cos 0 =
sin 0 =
(8.164a)
1
v2
c2
(8.164b)
1
1
v2
c2
1,
v2
1,
c2
(8.165)
v2 1
=
c2
(8.166)
136
E LECTROMAGNETIC R ADIATION
Hence, synchrotron radiation from ultra-relativistic charges is characterized by a radiation lobe width which is approximately
(8.167)
This angular interval is swept by the charge during the time interval
t0 =
(8.168)
(8.169)
in the direction toward the observer who therefore measures a pulse width of
length
l
vt0
v
v
v 1
= t0
= 1 t0 = 1
1
c
c
c
c 0
c 0
2
1 vc # 1 + vc # 1
1
v
1 1
"v
1 2
= 3
= "
c
2
2
0
* +-, .
0
0
1+
* +/, c.
1/2
2
(8.170)
t = t0
1
= 23 0
t
(8.171)
8.5
R ADIATION
137
2. The electrons are perfectly evenly distributed in the orbit. This is the
case, for instance, for electrons in a circular current in a conductor. In
this case the radiation fields cancel completely and no far fields are generated.
3. The electrons are unevenly distributed in the orbit. This happens
for an
(x x0 )
|x x0 | v
c
v
(8.172)
Integration over the solid angle gives the totally radiated power as
2
2
v
dU rad 0 q0 2 v 2 1 c2 sin
=
3
dt0
6c
1 vc22
(8.173)
v2
q0
1
(x x0 ) x 3
E = Ez =
40 s3
c2
b
q0
=
40 2 (vt)2 + b2 /2 3/2
(8.174)
which represents a contracted field, approaching the field of a plane wave. The
passage of this field pulse corresponds to a frequency distribution of the field
energy. Fourier transforming, we obtain
1
E, =
2
E (t) eit dt =
q
42 0 bv
b
K1
v
b
v
(8.175)
138
E LECTROMAGNETIC R ADIATION
vt
q0
v = v x
B
E z
F IGURE 8.8: The perpendicular field of a charge q0 moving with velocity
v = v x is E z .
Here, K1 is the Kelvin function (Bessel function of the second kind with imaginary argument) which behaves in such a way for small and large arguments
that
q
, b v
(8.176a)
E, 2
4 0 bv
E, 0, b v
(8.176b)
showing that the pulse length is of the order b/v.
Due to the equipartition of the field energy into the electric and magnetic
fields, the total field energy can be written
U = 0
2 3 0
E
d x = 0
bmax
bmin
2
E
vdt 2b db
(8.177)
where the volume integration is over the plane perpendicular to v. With the
use of Parsevals identity for Fourier transforms, formula (8.12) on page 101,
we can rewrite this as
U=
bmax
U d = 40 v
q2
22 0 v
v/
bmin
bmin
db
d
b
2
E, d 2b db
(8.178)
q2
v
ln
U 2
2 0 v
bmin
(8.179)
8.5
R ADIATION
139
2
c
d
ln
N d
bmin
(8.180)
d
c
2
ln
p1 p01
(8.181)
2
N d
ln
E1 cp1 cp01
m0 c2 E1 E10
(8.182)
(8.183)
B = (t, x)H = m 0 H
(8.184)
E+ B = 0
t
B = 0
H D = j(t, x)
t
(8.185a)
(8.185b)
(8.185c)
(8.185d)
140
E LECTROMAGNETIC R ADIATION
Assuming for simplicity that the electric permittivity and the magnetic
permeability , and hence the relative permittivity and the relative permeability m all have fixed values, independent on time and space, for each type
of material we consider, we can derive the general telegraphers equation [cf.
Equation (2.26) on page 27]
2 E
E
2 E
=0
2
t
t2
(8.186)
=0
2
t2
(8.187)
0 m 0
m
(8.188)
i = 1, 2, 3
(8.189)
(8.190)
k k k = kv =
v
v v
(8.191)
8.5
R ADIATION
141
(8.192)
where now k is the wave vector in the medium given by Equation (8.191) on the
preceding page. With these definitions, the vacuum formula for the associated
magnetic field, Equation (2.33) on page 28,
1
1
kE = kE
B = k E =
v
(8.193)
c
=
n k
(8.194)
where, in the last step, we used Equation (8.191) on the preceding page.
If the medium has a refractive index which, as is usually the case, dependent on frequency , we say that the medium is dispersive. Because in this case
also k() and (k), so that the group velocity
vg =
(8.195)
has a unique value for each frequency component, and is different from v .
Except in regions of anomalous dispersion, v is always smaller than c. In
a gas of free charges, such as a plasma, the refractive index is given by the
expression
2p
2
(8.196)
N q2
0 m
(8.197)
n2 () = 1
where
2p =
142
E LECTROMAGNETIC R ADIATION
is the plasma frequency. Here m and N denote the mass and number density,
respectively, of charged particle species . In an inhomogeneous plasma, N =
N (x) so that the refractive index and also the phase and group velocities are
space dependent. As can be easily seen, for each given frequency, the phase
and group velocities in a plasma are different from each other. If the frequency
is such that it coincides with p at some point in the medium, then at that
point v while vg 0 and the wave Fourier component at is reflected
there.
Vavilov-Cerenkov
radiation
As we saw in Subsection 8.1, a charge in uniform, rectilinear motion in vacuum does not give rise to any radiation; see in particular Equation (8.98a) on
page 122. Let us now consider a charge in uniform, rectilinear motion in a
medium with electric properties which are different from those of a (classical)
vacuum. Specifically, consider a medium where
= Const > 0
(8.198a)
= 0
(8.198b)
1
c
=
<c
n
0
(8.199)
Hence, in this particular medium, the speed of propagation of (the phase planes
of) electromagnetic waves is less than the speed of light in vacuum, which we
know is an absolute limit for the motion of anything, including particles. A
medium of this kind has the interesting property that particles, entering into
the medium at high speeds |v|, which, of course, are below the phase speed
in vacuum, can experience that the particle speeds are higher than the phase
8.5
R ADIATION
143
A(t, x) =
1
q0
1 q0
=
40 |x x0 | n (xx0 )v 40 s
c
1
1 q0 v
q0 v
=
40 c2 |x x0 | n (xx0 )v 40 c2 s
c
(8.200a)
(8.200b)
where now
(x x0) v
0
s = xx n
c
(8.201)
The need for the absolute value of the expression for s is obvious in the case
when v/c 1/n because then the second term can be larger than the first term;
if v/c 1/n we recover the well-known vacuum case but with modified phase
speed. We also note that the retarded and advanced times in the medium are
[cf. Equation (3.34) on page 41]
0
0
tret
= tret
(t, x x0 ) = t
k |x x0 |
|x x0 | n
= t
c
0
k
x
x0 | n
|
|x
|x
0
0
tadv
= tadv
(t, x x0 ) = t +
= t+
(8.202a)
(8.202b)
so that the usual time interval t t 0 between the time measured at the point of
observation and the retarded time in a medium becomes
t t0 =
|x x0 | n
c
(8.203)
For v/c 1/n, the retarded distance s, and therefore the denominators in
Equations (8.200) vanish when
n(x x0 )
nv
v
= x x0
cos c = x x0
c
c
(8.204)
c
nv
(8.205)
In the direction defined by this angle c , the potentials become singular. During
the time interval t t0 given by expression (8.203) above, the field exists within
a sphere of radius |x x0 | around the particle while the particle moves a distance
l = v(t t0 )
(8.206)
144
E LECTROMAGNETIC R ADIATION
q0
c
nv
(8.207)
The Vavilov-Cerenkov
cone is similar in nature to the Mach cone in acoustics.
8.5
R ADIATION
145
(8.208)
q0 ix0 /v 0
e
(y )(z0 ) x 1
2
(8.209)
This Fourier component can be used in the formulae derived for a linear current
in Subsection 8.3.1 if only we make the replacements
0 = n 2 0
n
k
c
(8.210a)
(8.210b)
In this manner, using j from Equation (8.209), the resulting Fourier trans
forms of the Vavilov-Cerenkov
magnetic and electric radiation fields can be
calculated from the expressions (8.3a) and (8.3b) on page 98, respectively.
The total energy content is then obtained from Equation (8.12) on page 101
(integrated over a closed sphere at large distances). For a Fourier component
one obtains [cf. Equation (8.15) on page 101]
Urad d
40 nc
(j k)e
ikx0 3 0
V
q0 2 n2
=
exp i
163 0 c3
dx
2
2
x0
kx0 cos dx0 sin2 d
v
(8.211)
where is the angle between the direction of motion, x 01 , and the direction to
The integral in (8.211) is singular of a Dirac delta type. If
the observer, k.
we limit the spatial extent of the motion of the particle to the closed interval
[X, X] on the x0 axis we can evaluate the integral to obtain
Urad d =
# X
q0 2 n2 sin2 sin2 1 nv
c cos v
"
d
2
43 0 c3
1 nvc cos # v
(8.212)
"
146
E LECTROMAGNETIC R ADIATION
Urad sin d = 2
q0 2 n2 sin2 c
22 0 c3
sin2
Urad d *( cos
+-, ).
& 1 + nvc
X
v
2
'
(8.213)
d
sin c
1
1
sin2
& 1 + nvc
X
v
2
'
d 1
c2
cX
n2 v2 n
sin2 x
dx
2
x
(8.214)
c2
cX
1 2 2
n
n v
leading to the final approximate result for the total energy loss in the frequency
interval (, + d)
q0 2 X
U rad d =
20 c2
1
c2
d
n2 v2
(8.215)
1
c2
d
n2 ()v2
(8.216)
This result was derived under the assumption that v/c > 1/n(), i.e., under the
condition that the expression inside the parentheses in the right hand side is
positive. For all media it is true that n() 1 when , so there exist al
ways a highest frequency for which we can obtain Vavilov-Cerenkov
radiation
from a fast charge in a medium. Our derivation above for a fixed value of n is
valid for each individual Fourier component.
8.5
147
B IBLIOGRAPHY
Bibliography
[1] R. B ECKER, Electromagnetic Fields and Interactions, Dover Publications, Inc.,
New York, NY, 1982, ISBN 0-486-64290-9.
[2] V. L. G INZBURG, Applications of Electrodynamics in Theoretical Physics
and Astrophysics, Revised third ed., Gordon and Breach Science Publishers,
New York, London, Paris, Montreux, Tokyo and Melbourne, 1989, ISBN 288124-719-9.
[3] J. D. JACKSON, Classical Electrodynamics, third ed., Wiley & Sons, Inc.,
New York, NY . . . , 1999, ISBN 0-471-30932-X.
[4] J. B. M ARION , AND M. A. H EALD, Classical Electromagnetic Radiation,
second ed., Academic Press, Inc. (London) Ltd., Orlando, . . . , 1980, ISBN 012-472257-1.
[5] W. K. H. PANOFSKY, AND M. P HILLIPS, Classical Electricity and Magnetism,
second ed., Addison-Wesley Publishing Company, Inc., Reading, MA . . . , 1962,
ISBN 0-201-05702-6.
[6] J. VANDERLINDE, Classical Electromagnetic Theory, John Wiley & Sons, Inc.,
New York, Chichester, Brisbane, Toronto, and Singapore, 1993, ISBN 0-47157269-1.
148
E LECTROMAGNETIC R ADIATION
F
Formulae
(F.1)
B = 0
(F.2)
E =
(F.3)
B
t
H = j+ D
t
(F.4)
Constitutive relations
D = E
B
H=
(F.5)
j = E
(F.7)
P = 0 E
(F.8)
(F.6)
(F.9)
A
t
(F.10)
149
150
F ORMULAE
1
=0
c2 t
(F.11)
(F.12)
(F.13)
k E
c
(F.14)
40 c |x|
V0
Brad
(x) =
(F.15)
(F.16)
(F.17)
(F.18)
F.3
151
S PECIAL R ELATIVITY
40 c |x|
Brad
(x) =
(F.19)
(F.20)
Brad
(x) =
(F.21)
(F.22)
q
v2
(x x0 ) v
E(t, x) =
(x
x
)
1
+ (x x0 )
0
3
2
40 s
c
c2
E(t, x)
B(t, x) = (x x0 )
c|x x0 |
s = x x0 (x x0 )
v
c
x x0 = (x x0 ) |x x0 |
t0
t
=
x
(F.23)
(F.24)
(F.25)
v
c
(F.26)
|x x0 |
s
(F.27)
q
v2
E(t, x) =
(x x0 ) 1 2
3
40 s
c
v E(t, x)
B(t, x) =
c2
s=
|x x0 |
(x x0 ) v
c
(F.28)
(F.29)
(F.30)
152
F ORMULAE
g =
6-77
1 0
0
0
0 1 0
0
0 0 1 0 8
0 0
0 1
3445
(F.31)
(F.32)
(F.33)
6-77
v
c
3445
0 0
0 0
0
0
1 08
0
0
0 1
1
1 2
(F.34)
(F.35)
(F.36)
dt
= c d
(F.37)
F.3.5 Four-velocity
u =
dx
v
= ,
ds
c
(F.38)
F.4
153
V ECTOR R ELATIONS
F.3.6 Four-momentum
p = m0 c2 u = (E, cp)
(F.39)
v
c
(F.40)
F.3.8 Four-potential
A = (, cA)
(F.41)
6-77
F =
A A
=
x x
3445
0 E x Ey Ez
Ex
0
cBz cBy
Ey cBz
0
cB x 8
Ez cBy cB x
0
(F.42)
x i
i=1
def
def
x i
xi
xi
(F.43)
i = i =
,
,
=
x1 x2 x3
, ,
x y z
(F.44)
154
F ORMULAE
(F.45a)
(F.45b)
= sin x 1 + cos x 2
(F.45c)
(F.46a)
x 3 = cos r sin
(F.46b)
(F.46c)
(F.47)
(F.48)
(F.49)
Volume element
d3x = dV = drdS = r2 dr d
(F.50)
(F.51)
a b = b a = i jk a j bk x i
(F.52)
a (b c) = (a b) c
(F.53)
F.4
155
V ECTOR R ELATIONS
a (b c) = b(a c) c(a b)
(F.54)
a (b c) + b (c a) + c (a b) = 0
(F.55)
(F.56)
(a b) (c d) = (a b d)c (a b c)d
(F.57)
() = +
(F.58)
(a) = a + a
(F.59)
(a) = a a
(F.60)
(a b) = b ( a) a ( b)
(F.61)
(a b) = a( b) b( a) + (b )a (a )b
(F.62)
(a b) = a ( b) + b ( a) + (b )a + (a )b
(F.63)
= 2
(F.64)
= 0
(F.65)
( a) = 0
(F.66)
( a) = ( a) 2 a
(F.67)
156
F ORMULAE
Special relations
In the following k is an arbitrary constant vector.
x = 3
(F.68)
x = 0
(F.69)
x
|x|
(F.70)
|x| =
1
x
= 3
|x|
|x|
(F.71)
= 2
|x|3
1
= 4(x)
|x|
(F.72)
k
1
kx
= k
= 3
|x|
|x|
|x|
(F.73)
x
kx
=
k
|x|3
|x|3
k
= k2
|x|
if |x| 6= 0
1
= 4k(x)
|x|
(F.75)
(k a) = k( a) + k ( a) (k a)
(F.74)
(F.76)
F.4
157
B IBLIOGRAPHY
Integral relations
Let V(S ) be the volume bounded by the closed surface S (V). Denote the 3dimensional volume element by d3x( dV) and the surface element, directed
along the outward pointing surface normal unit vector n,
by dS( d2x n).
( a) d3x =
() d3x =
dS a
dS
(F.77)
(F.78)
( a) d3x =
dS a
(F.79)
If S (C) is an open surface bounded by the contour C(S ), whose line element is dl, then
dl =
a dl =
dS
dS ( a)
(F.80)
(F.81)
Bibliography
[1] George B. Arfken and Hans J. Weber. Mathematical Methods for Physicists. Academic Press, Inc., San Diego, CA . . . , fourth, international edition, 1995. ISBN 0-12-059816-7.
[2] Philip M. Morse and Herman Feshbach. Methods of Theoretical Physics. Part I. McGraw-Hill Book Company, Inc., New York, NY . . . , 1953.
ISBN 07-043316-8.
[3] Wolfgang K. H. Panofsky and Melba Phillips. Classical Electricity and
Magnetism. Addison-Wesley Publishing Company, Inc., Reading, MA . . . ,
second edition, 1962. ISBN 0-201-05702-6.
158
F ORMULAE
M
Mathematical
Methods
M.1.1 Vectors
Radius vector
Any vector can be represented mathematically in several different ways. One
suitable representation is in terms of an ordered N-tuple, or row vector, of the
coordinates xN where N is the dimensionality of the space under consideration.
The most basic vector is radius vector which is the vector from the origin to the
point of interest. Its N-tuple representation simply enumerates the coordinates
159
160
M ATHEMATICAL M ETHODS
which describe this point. In this sense, the radius vector from the origin to a
point is synonymous with the coordinates of the point itself.
In the 3D space 3 , we have N = 3 and the radius vector can be represented
by the triplet (x1 , x2 , x3 ) of coordinates xi , i = 1, 2, 3. The coordinates xi are
scalar quantities which describe the position along the unit base vectors x i
which span 3 . Therefore a representation of the radius vector in 3 is
3
x = x i xi x i xi
def
(M.1)
i=1
xi (x1 , x2 , x3 ) (x, y, z)
(M.2)
x (x0 , x1 , x2 , x3 )
(M.3)
x (x0 , x1 , x2 , x3 )
(M.4)
M.1
S CALARS , V ECTORS
AND
161
T ENSORS
M.1.2 Fields
A field is a physical entity which depends on one or more continuous parameters. Such a parameter can be viewed as a continuous index which enumerates the coordinates of the field. In particular, in a field which depends on
the usual radius vector x of 3 , each point in this space can be considered as
one degree of freedom so that a field is a representation of a physical entity
which has an infinite number of degrees of freedom.
Scalar fields
We denote an arbitrary scalar field in
by
def
(M.6)
(x0 , x1 , x2 , x3 ) (x )
(M.7)
which indicates that the four-scalar depends on all four coordinates spanning
this space. Since a four-scalar has the same value at a given point regardless
of coordinate system, it is also called an invariant.
Analogous to the transformation rule, Equation (M.5) on the facing page,
for the differential dx , the transformation rule for the differential operator
/x under a transformation x x0 becomes
x
=
x0 x0 x
(M.8)
as follows:
a(x) = x i ai (x)
(M.9)
(M.10)
162
M ATHEMATICAL M ETHODS
(M.11)
(M.12)
x0
y
x
(M.13)
x
y
x0
(M.14)
A(x) =
53
Ai j (xk ) #
def
"
(M.15)
where, in the last member, we have again used the more compact component
notation. Strictly speaking, the tensor field described here is a tensor of rank
two.
M.1
S CALARS , V ECTORS
AND
163
T ENSORS
0 if i 6= j
1 if i = j
(M.16)
53
(i j ) =
1 0 0
0 1 08
0 0 1
(M.17)
i jk =
n
mo
lm
1
0
1
(M.18)
(M.19)
In fact, tensors may have any rank n. In this picture a scalar is considered
to be a tensor of rank n = 0 and a vector a tensor of rank n = 1. Consequently,
the notation where a vector (tensor) is represented in its component form is
called the tensor notation. A tensor of rank n = 2 may be represented by a twodimensional array or matrix whereas higher rank tensors are best represented
in their component forms (tensor notation).
In 4D, we have three forms of four-tensor fields of rank n. We speak of
a contravariant four-tensor field, denoted A1 2 ...n (x ),
a covariant four-tensor field, denoted A1 2 ...n (x ),
...
a (x ) g a (x )
(M.20)
164
M ATHEMATICAL M ETHODS
This rule is often called lowering of index. The raising of index analogue of
the index lowering rule is:
def
a (x ) g a (x )
(M.21)
More generally, the following lowering and raising rules hold for arbitrary
rank n mixed tensor fields:
2 ...k1 k
1 2 ...k1
gk k A1k+1
k+2 ...n (x ) = Ak k+1 ...n (x )
(M.22)
1 2 ...k1 k
2 ...k1
gk k A1k k+1
...n (x ) = Ak+1 k+2 ...n (x )
(M.23)
Successive lowering and raising of more than one index is achieved by a repeated application of this rule. For example, a dual application of the lowering
operation on a rank 2 tensor in contravariant form yields
A = g g A
(M.24)
i.e., the same rank 2 tensor in covariant form. This operation is also known as
a tensor contraction.
T ENSORS IN 3D SPACE
E XAMPLE
M.1
d2x
F IGURE M.1: Surface element d2x and the unit normal vector n.
Consider the tetrahedron-like volume element V indicated in Figure M.1 on the next
page of a solid, fluid, or gaseous body, whose atomistic structure is irrelevant for the
present analysis. Let dS = d2x n in Figure M.1 on the facing page be a directed surface
element of this volume element and let the vector T n d2x be the force that matter, lying
on the side of d2x toward which the unit normal vector n points, acts on matter which
lies on the opposite side of d2x. This force concept is meaningful only if the forces
are short-range enough that they can be assumed to act only in the surface proper.
According to Newtons third law, this surface force fulfils
T n = T n
(M.25)
Using (M.25) and Newtons second law, we find that the matter of mass m, which at a
M.1
S CALARS , V ECTORS
AND
165
T ENSORS
x3
d2x
x2
V
x1
F IGURE M.2: Terahedron-like volume element V containing matter.
given instant is located in V obeys the equation of motion
T n d2x T x 1 d2x T x 2 d2x T x 3 d2x + Fext = ma
(M.26)
where Fext is the external force and a is the acceleration of the volume element. In
other words
m
Fext
T n = n1 T x 1 + n2 T x 2 + n3 T x 3 + 2 a
(M.27)
dx
m
Since both a and Fext /m remain finite whereas m/d2x 0 as V 0, one finds that in
this limit
3
T n = ni T x i ni T x i
(M.28)
i=1
From the above derivation it is clear that Equation (M.28) above is valid not only in
equilibrium but also when the matter in V is in motion.
Introducing the notation
T i j = = T x i >
(M.29)
for the jth component of the vector T x i , we can write Equation (M.28) on the preced-
166
M ATHEMATICAL M ETHODS
T n j = (T n ) j = ni T i j ni T i j
(M.30)
i=1
Using Equation (M.30) above, we find that the component of the vector T n in the
direction of an arbitrary unit vector m
is
T n m = T n m
= T n j m j =
j=1
j=1
ni T i j i
(M.31)
m j ni T i j m j = n
T m
i=1
Hence, the jth component of the vector T x i , here denoted T i j , can be interpreted as the
i jth component of a tensor T. Note that T n m is independent of the particular coordinate
system used in the derivation.
We shall now show how one can use the momentum law (force equation) to derive
the equation of motion for an arbitrary element of mass in the body. To this end we
consider a part V of the body. If the external force density (force per unit volume) is
denoted by f and the velocity for a mass element dm is denoted by v, we obtain
d
dt f
v dm =
V
f V f d x + f S Tn d x
(M.32)
f V dt v j dm = f V f j d x + f S T nj d x = f V f j d x + f S ni Ti j d x
(M.33)
where, in the last step, Equation (M.30) was used. Setting dm = d 3x and using the
divergence theorem on the last term, we can rewrite the result as
fV
d
v j d3x =
f
dt
f j d3x +
fV
T i j 3
dx
xi
(M.34)
Since this formula is valid for any arbitrary volume, we must require that
T i j
d
vj fj
=0
dt
xi
(M.35)
or, equivalently
v j
T i j
+ v v j f j
=0
t
xi
(M.36)
Note that v j /t is the rate of change with time of the velocity component v j at a fixed
point x = (x1 , x1 , x3 ).
E ND OF
EXAMPLE
M.1
M.1
S CALARS , V ECTORS
AND
167
T ENSORS
(M.37)
(M.38)
where we made use of the index lowering and raising rules (M.20) and (M.21).
The result is a four-scalar, i.e., an invariant which is independent on in which
inertial system it is measured.
The quadratic differential form
ds2 = g dx dx = dx dx
(M.39)
i.e., the scalar product of the differential radius four-vector with itself, is an
invariant called the metric. It is also the square of the line element ds which is
the distance between neighbouring points with coordinates x and x + dx .
I NNER PRODUCTS IN
def
6 ),
expressed in terms of
E XAMPLE
M.2
(M.40)
def
(M.41)
(M.42)
Using this in Equation (M.40), we see that we can interpret this so that the complex
168
M ATHEMATICAL M ETHODS
unit vector is
A
A = =
A
aR
a R + i
aI
a R + i
a I
a I
(M.43)
On the other hand, the definition of the scalar product in terms of the inner product of
complex vector with its own complex conjugate yields
def
(M.44)
aR
a2R + a2I
a R + i
aR p a2R + a2I
a2R + a2I
aI
a R + i
a2R + a2I
a I
aI p a2R + a2I
a2R + a2I
(M.45)
a I
E ND OF
E XAMPLE
M.3
METRIC IN
EXAMPLE
M.2
L ORENTZ SPACE
In q 4 the metric tensor attains a simple form [see Equation (5.7) on page 54 for an
example] and, hence, the scalar product in Equation (M.38) on the preceding page can
be evaluated almost trivially and becomes
a b = (a0 , a) (b0 , b) = a0 b0 a b
(M.46)
4.
The
(M.47)
4
(M.48)
E ND OF
EXAMPLE
M.3
M.1
S CALARS , V ECTORS
AND
169
T ENSORS
IJJK
g =
e
0
0
0
0
e
0
0
0
0
r2
0
LNMM
0
0
O
0
2
2
r sin
E XAMPLE
M.4
(M.49)
where = (ct, r, , ) and = (ct, r, , ). In such a space, the metric takes the form
ds2 = c2 e (dt)2 e (dr)2 r2 (d)2 r2 sin2 (d)2
(M.50)
In general relativity the metric tensor is not given a priori but is determined by the
Einstein equations.
E ND OF
EXAMPLE
M.4
Dyadic product
The dyadic product field A(x) a(x)b(x) with two juxtaposed vector fields
a(x) and b(x) is the outer product of a and b. Operating on this dyad from the
right and from the left with an inner product of an vector c one obtains
def
def
def
def
A c ab c a(b c)
(M.51a)
c A c ab (c a)b
(M.51b)
i.e., new vectors, proportional to a and b, respectively. In mathematics, a dyadic product is often called tensor product and is frequently denoted a b.
In matrix notation the outer product of a and b is written
"
ab = x 1
x 2
x 3 #
53
a1 b1 a1 b2 a1 b3
a1 b2 a2 b2 a2 b3 8
a1 b3 a3 b2 a3 b3
53
x 1
x 2 8
x 3
(M.52)
which means that we can represent the tensor A(x) in matrix form as
" Ai j (xk )
53
a1 b1 a1 b2 a1 b3
a1 b2 a2 b2 a2 b3 8
a1 b3 a3 b2 a3 b3
(M.53)
which we identify with expression (M.15) on page 162, viz. a tensor in matrix
notation.
170
M ATHEMATICAL M ETHODS
Vector product
The vector product or cross product of two arbitrary 3D vectors a and b in
ordinary 3 space is the vector
c = a b = i jk a j bk x i
(M.54)
x i
def
xi
(M.55)
where x i is the ith unit vector in a Cartesian coordinate system. Since the
operator in itself has vectorial properties, we denote it with a boldface nabla.
In component notation we can write
i =
,
,
x1 x2 x3
(M.56)
M.1
S CALARS , V ECTORS
AND
171
T ENSORS
,
,
,
x0 x1 x2 x3
(M.57)
,
,
,
x0 x1 x2 x3
(M.58)
(M.59)
y0 = 0 x # y
(M.60)
"
and
"
respectively.
In q
1
,
c t
(M.61)
E XAMPLE
M.5
1
, =
c t
1
,
c t
(M.62)
d
1 2
2 =
2
2
c t
(M.63)
, and sometimes defined with
E ND OF
EXAMPLE
M.5
172
M ATHEMATICAL M ETHODS
With the help of the del operator we can define the gradient, divergence
and curl of a tensor (in the generalised sense).
The gradient
The gradient of an
a(x):
vector field
(M.64)
From this we see that the boldface notation for the nabla and del operators is
very handy as it elucidates the 3D vectorial property of the gradient.
In 4D, the four-gradient is a covariant vector, formed as a derivative of a
four-scalar field (x ), with the following component form:
(x ) =
E XAMPLE
M.6
G RADIENTS OF
(x )
x
(M.65)
3D
= 0
x0i
(M.66)
Using this, the unprimed version, Equation (M.55) on page 170, and elementary
rules of differentiation, we obtain the following two very useful results:
= |x x0 | > = x i
|x x0 |
x x0
|x x0 |
=
=
i
xi
|x x0 |
x0i
(M.67)
= 0 = |x x0 | >
and
x x0
1
=
= 0
0
|x x0 |3
|x x |
1
|x x0 |
(M.68)
E ND OF
EXAMPLE
M.6
M.1
S CALARS , V ECTORS
AND
173
T ENSORS
The divergence
We define the 3D divergence of a vector field in
a(x) = x j a j (x) = i j i a j (x) = i ai (x) =
as
ai (x)
= (x)
xi
(M.69)
a (x )
x
(M.70)
D IVERGENCE IN 3D
For an arbitrary
0 a(x0 )
a(x0 )
=
0
+ a(x0 ) 0
|x x0 |
|x x0 |
1
|x x0 |
(M.71)
E XAMPLE
M.7
which demonstrates how the primed divergence, defined in terms of the primed
del operator in Equation (M.66) on the preceding page, works.
E ND OF
EXAMPLE
M.7
The Laplacian
The 3D Laplace operator or Laplacian can be described as the divergence of
the gradient operator:
2 = = =
2
2
x i x j
= i j i j = 2i = 2 2
xi
x j
xi i=1 xi
(M.72)
The symbol 2 is sometimes read del squared. If, for a scalar field (x),
2 < 0 at some point in 3D space, it is a sign of concentration of at that
point.
174
M ATHEMATICAL M ETHODS
E XAMPLE
M.8
T HE L APLACIAN AND
D IRAC DELTA
THE
1
= 2
|x x0 |
is
1
= 4(x x0 )
|x x0 |
(M.73)
E ND OF
EXAMPLE
M.8
The curl
In 3 the curl of a vector field a(x), denoted a(x), is another
b(x) which can be defined in the following way:
a(x) = i jk x i j ak (x) = i jk x i
ak (x)
= b(x)
x j
vector field
(M.74)
where use was made of the Levi-Civita tensor, introduced in Equation (M.18)
on page 163.
The covariant 4D generalisation of the curl of a four-vector field a (x ) is
the antisymmetric four-tensor field
G (x ) = a (x ) a (x ) = G (x )
(M.75)
E XAMPLE
M.9
T HE
CURL OF A GRADIENT
Using the definition of the 3 curl, Equation (M.74) above, and the gradient, Equation (M.64) on page 172, we see that
[(x)] = i jk x i j k (x)
(M.76)
2
2
(x) x 1
x2 x3 x3 x2
2
2
(x) x 2
x3 x1 x1 x3
+
+
(x) x i
x j xk
(M.77)
2
2
(x) x 3
x1 x2 x2 x1
M.1
S CALARS , V ECTORS
AND
175
T ENSORS
(M.78)
(M.79)
3.
T HE DIVERGENCE OF
EXAMPLE
M.9
A CURL
With the use of the definitions of the divergence (M.69) and the curl, Equation (M.74)
on the preceding page, we find that
[ a(x)] = i [ a(x)]i = i jk i j ak (x)
(M.80)
E XAMPLE
M.10
Using the definition for the Levi-Civita symbol, defined by Equation (M.18) on
page 163, we find that, due to the assumed well-behavedness of a(x),
i i jk j ak (x) =
=
i jk
ak
xi
x j
2
2
a1 (x)
x2 x3 x3 x2
2
2
a2 (x)
x3 x1 x1 x3
+
+
(M.81)
2
2
a3 (x)
x1 x2 x2 x1
0
i.e., that
[ a(x)] 0
(M.82)
a (x ) 6= 0
(M.83)
E ND OF EXAMPLE M.10
176
M ATHEMATICAL M ETHODS
L(qi , q i , t) = L qi ,
dqi
,t = T V
dt
(M.84)
where qi is the generalised coordinate, T the kinetic energy and V the potential energy of a mechanical system, The Lagrangian satisfies the Lagrange
equations
L
L
=0
q i
qi
(M.85)
L
q i
(M.86)
(M.87)
(M.88)
After differentiating the left and right hand sides of this definition and setting
them equal we obtain
H
H
H
L
L
L
dpi +
dqi +
dt = q i dpi + pi dq i
dqi
dq i dt (M.89)
pi
qi
t
qi
q i
t
M.2
177
B IBLIOGRAPHY
(M.90a)
(M.90b)
Bibliography
[1] George B. Arfken and Hans J. Weber. Mathematical Methods for Physicists. Academic Press, Inc., San Diego, CA . . . , fourth, international edition, 1995. ISBN 0-12-059816-7.
[2] R. A. Dean. Elements of Abstract Algebra.
New York, NY . . . , 1967. ISBN 0-471-20452-8.
[3] Arthur A. Evett. Permutation symbol approach to elementary vector analysis. American Journal of Physics, 34, 1965.
[4] Philip M. Morse and Herman Feshbach. Methods of Theoretical Physics. Part I. McGraw-Hill Book Company, Inc., New York, NY . . . , 1953.
ISBN 07-043316-8.
[5] Barry Spain. Tensor Calculus. Oliver and Boyd, Ltd., Edinburgh and
London, third edition, 1965. ISBN 05-001331-9.
178
M ATHEMATICAL M ETHODS
Index
179
conservative field, 11
conservative forces, 77
constitutive relations, 14
contraction, 54
contravariant component form, 54,
160
contravariant field tensor, 67
contravariant four-tensor field, 163
contravariant four-vector, 162
contravariant four-vector field, 57
contravariant vector, 54
convection potential, 125
convective derivative, 11
cosine integral, 104
Coulomb gauge, 38
Coulombs law, 2
covariant, 51
covariant component form, 160
covariant field tensor, 66
covariant four-tensor field, 163
covariant four-vector, 162
covariant four-vector field, 57
covariant vector, 54
cross product, 170
curl, 174
cutoff, 131
cyclotron radiation, 134, 137
dAlembert operator, 36, 62, 171
del operator, 170
del squared, 173
differential distance, 56
differential vector operator, 170
180
M ATHEMATICAL M ETHODS
181
182
M ATHEMATICAL M ETHODS
183
polarisation currents, 90
polarisation potential, 105
polarisation vector, 105
positive definite, 58
positive definite norm, 55
potential energy, 77, 176
potential theory, 106
power flux, 93
Poynting vector, 93
Poyntings theorem, 92
Proca Lagrangian, 85
propagator, 39
proper time, 56
pseudoscalar, 159
pseudoscalars, 170
pseudotensor, 159
pseudotensors, 170
pseudovector, 65, 159, 170
quadratic differential form, 56, 167
quantum mechanical nonlinearity, 3
radiation field, 46, 49, 118
radiation fields, 97
radiation gauge, 38
radiation resistance, 103
radius four-vector, 53
radius vector, 159
raising of index, 164
rank, 162
rapidity, 60
refractive index, 140
relative electric permittivity, 94
relative magnetic permeability, 94
relative permeability, 140
relative permittivity, 140
Relativity principle, 51
relaxation time, 25
rest mass density, 81
retarded Coulomb field, 49
retarded potentials, 41
retarded relative distance, 113
retarded time, 41
Riemannian metric, 56
Riemannian space, 54, 160
row vector, 159
scalar, 159, 173
scalar field, 57, 161
scalar product, 167
shock front, 144
signature, 54
simultaneous coordinate, 121
skew-symmetric, 66
skin depth, 30
source point, 3
space components, 55
space-like interval, 56
space-time, 55
special theory of relativity, 51
spherical Bessel function of the first
kind, 106
spherical Hankel function of the first
kind, 106
spherical waves, 99
super-potential, 105
synchrotron radiation, 134, 137
synchrotron radiation lobe width, 135
telegraphers equation, 27, 140
temporal dispersive media, 10
temporal gauge, 38
tensor, 159
tensor contraction, 164
tensor field, 162
tensor notation, 163
tensor product, 169
three-dimensional functional derivative, 79
time component, 55
184
M ATHEMATICAL M ETHODS
time-harmonic wave, 25
time-independent diffusion equation,
26
time-independent telegraphers equation, 29
time-independent wave equation, 26
time-like interval, 56
total charge, 87
transverse components, 27
transverse gauge, 38
vacuum permeability, 5
vacuum permittivity, 2
vacuum polarisation effects, 3
vacuum wave number, 26
Vavilov-Cerenkov
radiation, 142, 144
vector, 159
vector product, 170
velocity field, 118
virtual simultaneous coordinate, 114,
118
wave vector, 28, 140
world line, 58
Youngs modulus, 77
Yukawa meson field, 84